Limits...
Challenges of ascertaining national trends in the incidence of coronary heart disease in the United States.

Ford ES, Roger VL, Dunlay SM, Go AS, Rosamond WD - J Am Heart Assoc (2014)

View Article: PubMed Central - PubMed

Affiliation: Division of Population Health, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, GA (E.S.F.).

ABSTRACT

Although a complete picture of the national trend in CHD incidence in the United States remains elusive, the findings from community‐based studies, prospective studies, and health care delivery systems reporting decreases in incidence of CHD provide the most convincing evidence that the national incidence of CHD may have declined. These findings are buttressed by data showing declines in national rates of death attributed to CHD, studies showing decreases in sudden death and out‐of hospital mortality associated with MI, declines in hospitalizations for CHD, improving MI severity, possible recent declines in prevalence of CHD, declines in predicted 10‐year risk, and favorable improvements in the prevalence and control of some major CHD risk factors. Although each of these pieces of information is an imperfect reflection of incident CHD, in the aggregate they tell an increasingly compelling story of the evolution of CHD incidence in the United States. Declines in death from CHD are potentially suggestive of declining CHD incidence if the declines in case‐fatality rates do not account for the entire decrease in mortality.

Because the studies examining trends in CHD incidence covered different time frames and were conducted in different areas of the United States, pinpointing the exact time when incidence started to decrease is difficult because the onset of the start of any declines may have varied geographically. Community surveillance studies have reported decreases in incidence as early as the 1960s (Rochester, MN),7 during the late 1980s (Corpus Christi Heart Project),76 1990s (ARIC),38 and 2000s (Worcester Heart Attack Study).32 Other studies suggest that decreases in incidence occurred during the 1960s (Framingham Heart Study, the Du Pont Company),10,24 1970s (Framingham Heart Study),31 1980s (Nurses' Health Study, NHANES Epidemiologic Follow‐up Study, Framingham Heart Study),28,31,79 1990s (Framingham Heart Study, Nurses' Health Study),31,79 and 2000s (Kaiser Permanente Northern California).33

Three of the studies illustrate the difficulty in interpreting surveillance data over long periods of time particularly when changes in diagnostic criteria occur.31,34,38 The introduction of troponin testing around the turn of the century marked an important change in the diagnostic criteria for MI91 and coincided with a shift in the ratio of STEMI to NSTEMI with decreases in rates of STEMI and increases in rates of NSTEMI. Furthermore, the advent of electron‐beam computed tomography and multi‐detector computed tomography to detect calcium in the walls of coronary arteries has led to earlier identification of CHD.92 From a surveillance point of view, these disruptive changes in diagnostic criteria emphasize the importance of being able to disentangle the effects on such changes on trend analyses.

Validation of incident CHD events enhances the credibility of trends in CHD incidence. The majority of community surveillance, cohort, and health care delivery system‐based studies included reviews of medical records searching for clinical presentation, electrographic criteria, and cardiac biomarkers to confirm the presence of CHD, although these validation efforts differed across studies and across time periods as diagnostic criteria were also evolving.

Furthermore, observational studies suggest that an enormous amount of CHD can yet be prevented by adopting healthy behaviors or by optimizing behavioral and clinical risk factors as exemplified by the AHA's 7 cardiovascular health metrics.93–99 In addition, initiatives such as the Million Hearts Initiative, which aims to prevent 1 million heart attacks and strokes by 2017 through a combination of clinical and community actions, will, if successful, potentially hasten the decline in the incidence of CHD.100–101

The data sources opening a window into race or ethnicity‐specific trends of CHD incidence are few. Data from the ARIC study suggest that African‐American men and women did enjoy declining CHD incidence, but the decline among African Americans manifested itself later than among whites and the size of the decline was smaller than that of whites. These results are corroborated by Medicare data and data from the NIS also showing that the hospitalization rate for MI declined more slowly among African Americans than among whites.21–22 Gaps in evidence exist about the trends in CHD incidence among other racial or ethnic groups such as Hispanics and Asians. Given the rapidly evolving demographic composition of the US population, data collection efforts to shed light on the evolution of CHD in major and growing racial and ethnic groups are needed. Perhaps, large health care delivery systems and growing health system‐based networks are best suited to provide such results if their expanding electronic medical record and other data systems capture valid racial and ethnic designations and relevant clinical outcomes of their memberships.

Efforts to establish community surveillance for CHD harken back decades.102–103 A national system to monitor CHD incidence has never been established, however, and this gap has not gone unnoticed.104–107 As part of its recommendations, the Institute of Medicine highlighted the critical importance of having data on the incidence of CVD and the need for a system that would collect such data. The report cited potential avenues such as the establishment of registries, the use of cohort studies, and the use of claims and electronic medical record data to accomplish such a goal. The development of a national system to monitor the trend in the incidence of CHD would help to fill this current void in the knowledge base of the epidemiology of CHD and provide critical data to improve cardiovascular health of the US population.

In conclusion, definitive data about national trends of incident CHD in the United States currently are not available, and, therefore, clues about these trends must be gleaned from a variety of auxiliary data sources. Studies in different parts of the country demonstrate improvements in the incidence of CHD that may have commenced several decades ago in some parts of the country, and an increasing number of recent studies have described favorable trends during the first decade of the 21st century. Taken together, these studies yield encouraging but tentative signals that the incidence of CHD in the United States may be waning. Bringing greater clarity to this important topic of cardiovascular epidemiology poses a pressing public health need.

Show MeSH

Related in: MedlinePlus

Age‐adjusted mortality rates from CHD for adults aged ≥25 years, United States. Results were generated with WONDER using the Compressed Mortality File of the National Vital Statistics System. For the period 1979–1999, International Classification of Diseases 9 codes 410‐414 and 429.2 were used. For 2000–2009, International Classification of Diseases codes I20‐O25 were used. Results were age‐adjusted to the projected year 2000 US population. CHD indicates coronary heart disease.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4338697&req=5

fig01: Age‐adjusted mortality rates from CHD for adults aged ≥25 years, United States. Results were generated with WONDER using the Compressed Mortality File of the National Vital Statistics System. For the period 1979–1999, International Classification of Diseases 9 codes 410‐414 and 429.2 were used. For 2000–2009, International Classification of Diseases codes I20‐O25 were used. Results were age‐adjusted to the projected year 2000 US population. CHD indicates coronary heart disease.

Mentions: The category “diseases of the heart” has long been and continues to be the leading cause of death in the United States based on data from death certificates.4 After increasing during the first part of the 20th century, the mortality rate attributed to CHD peaked during the late 1960s and reversed course starting a prolonged and continuing decline.5–6 From 1980 through 2009, age‐adjusted CHD mortality has decreased by 66% among men and 67% among women (Figure 1). Furthermore, age‐adjusted rates decreased by 60% among African American women, 57% among African American men, 68% among white women, and 67% among white men (Figure 2). CHD mortality was defined as International Classification of Diseases (ICD)‐9 codes 410‐414 and 429.2 or ICD‐10 codes I20‐I25. Regional studies such as the Framingham Heart Study, the Minnesota Heart Survey, Honolulu Heart Program, and the Atherosclerosis Risk in Communities Study (ARIC) also described declining rates of CHD mortality.7–11 The factors contributing to the decline have been debated, and a combination of treatment and improvements in population levels of risk factors for CHD has been credited with lowering the CHD mortality rate.12–17 The declining mortality rates raised the prospect of declining incidence rates. Because mortality rates are subject to a number of influences such as disease severity, case fatality, changes in risk factors, improved treatment, and incident or new cases,18 declining mortality rates alone cannot automatically be equated with declining incidence rates.


Challenges of ascertaining national trends in the incidence of coronary heart disease in the United States.

Ford ES, Roger VL, Dunlay SM, Go AS, Rosamond WD - J Am Heart Assoc (2014)

Age‐adjusted mortality rates from CHD for adults aged ≥25 years, United States. Results were generated with WONDER using the Compressed Mortality File of the National Vital Statistics System. For the period 1979–1999, International Classification of Diseases 9 codes 410‐414 and 429.2 were used. For 2000–2009, International Classification of Diseases codes I20‐O25 were used. Results were age‐adjusted to the projected year 2000 US population. CHD indicates coronary heart disease.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4338697&req=5

fig01: Age‐adjusted mortality rates from CHD for adults aged ≥25 years, United States. Results were generated with WONDER using the Compressed Mortality File of the National Vital Statistics System. For the period 1979–1999, International Classification of Diseases 9 codes 410‐414 and 429.2 were used. For 2000–2009, International Classification of Diseases codes I20‐O25 were used. Results were age‐adjusted to the projected year 2000 US population. CHD indicates coronary heart disease.
Mentions: The category “diseases of the heart” has long been and continues to be the leading cause of death in the United States based on data from death certificates.4 After increasing during the first part of the 20th century, the mortality rate attributed to CHD peaked during the late 1960s and reversed course starting a prolonged and continuing decline.5–6 From 1980 through 2009, age‐adjusted CHD mortality has decreased by 66% among men and 67% among women (Figure 1). Furthermore, age‐adjusted rates decreased by 60% among African American women, 57% among African American men, 68% among white women, and 67% among white men (Figure 2). CHD mortality was defined as International Classification of Diseases (ICD)‐9 codes 410‐414 and 429.2 or ICD‐10 codes I20‐I25. Regional studies such as the Framingham Heart Study, the Minnesota Heart Survey, Honolulu Heart Program, and the Atherosclerosis Risk in Communities Study (ARIC) also described declining rates of CHD mortality.7–11 The factors contributing to the decline have been debated, and a combination of treatment and improvements in population levels of risk factors for CHD has been credited with lowering the CHD mortality rate.12–17 The declining mortality rates raised the prospect of declining incidence rates. Because mortality rates are subject to a number of influences such as disease severity, case fatality, changes in risk factors, improved treatment, and incident or new cases,18 declining mortality rates alone cannot automatically be equated with declining incidence rates.

View Article: PubMed Central - PubMed

Affiliation: Division of Population Health, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, GA (E.S.F.).

ABSTRACT

Although a complete picture of the national trend in CHD incidence in the United States remains elusive, the findings from community‐based studies, prospective studies, and health care delivery systems reporting decreases in incidence of CHD provide the most convincing evidence that the national incidence of CHD may have declined. These findings are buttressed by data showing declines in national rates of death attributed to CHD, studies showing decreases in sudden death and out‐of hospital mortality associated with MI, declines in hospitalizations for CHD, improving MI severity, possible recent declines in prevalence of CHD, declines in predicted 10‐year risk, and favorable improvements in the prevalence and control of some major CHD risk factors. Although each of these pieces of information is an imperfect reflection of incident CHD, in the aggregate they tell an increasingly compelling story of the evolution of CHD incidence in the United States. Declines in death from CHD are potentially suggestive of declining CHD incidence if the declines in case‐fatality rates do not account for the entire decrease in mortality.

Because the studies examining trends in CHD incidence covered different time frames and were conducted in different areas of the United States, pinpointing the exact time when incidence started to decrease is difficult because the onset of the start of any declines may have varied geographically. Community surveillance studies have reported decreases in incidence as early as the 1960s (Rochester, MN),7 during the late 1980s (Corpus Christi Heart Project),76 1990s (ARIC),38 and 2000s (Worcester Heart Attack Study).32 Other studies suggest that decreases in incidence occurred during the 1960s (Framingham Heart Study, the Du Pont Company),10,24 1970s (Framingham Heart Study),31 1980s (Nurses' Health Study, NHANES Epidemiologic Follow‐up Study, Framingham Heart Study),28,31,79 1990s (Framingham Heart Study, Nurses' Health Study),31,79 and 2000s (Kaiser Permanente Northern California).33

Three of the studies illustrate the difficulty in interpreting surveillance data over long periods of time particularly when changes in diagnostic criteria occur.31,34,38 The introduction of troponin testing around the turn of the century marked an important change in the diagnostic criteria for MI91 and coincided with a shift in the ratio of STEMI to NSTEMI with decreases in rates of STEMI and increases in rates of NSTEMI. Furthermore, the advent of electron‐beam computed tomography and multi‐detector computed tomography to detect calcium in the walls of coronary arteries has led to earlier identification of CHD.92 From a surveillance point of view, these disruptive changes in diagnostic criteria emphasize the importance of being able to disentangle the effects on such changes on trend analyses.

Validation of incident CHD events enhances the credibility of trends in CHD incidence. The majority of community surveillance, cohort, and health care delivery system‐based studies included reviews of medical records searching for clinical presentation, electrographic criteria, and cardiac biomarkers to confirm the presence of CHD, although these validation efforts differed across studies and across time periods as diagnostic criteria were also evolving.

Furthermore, observational studies suggest that an enormous amount of CHD can yet be prevented by adopting healthy behaviors or by optimizing behavioral and clinical risk factors as exemplified by the AHA's 7 cardiovascular health metrics.93–99 In addition, initiatives such as the Million Hearts Initiative, which aims to prevent 1 million heart attacks and strokes by 2017 through a combination of clinical and community actions, will, if successful, potentially hasten the decline in the incidence of CHD.100–101

The data sources opening a window into race or ethnicity‐specific trends of CHD incidence are few. Data from the ARIC study suggest that African‐American men and women did enjoy declining CHD incidence, but the decline among African Americans manifested itself later than among whites and the size of the decline was smaller than that of whites. These results are corroborated by Medicare data and data from the NIS also showing that the hospitalization rate for MI declined more slowly among African Americans than among whites.21–22 Gaps in evidence exist about the trends in CHD incidence among other racial or ethnic groups such as Hispanics and Asians. Given the rapidly evolving demographic composition of the US population, data collection efforts to shed light on the evolution of CHD in major and growing racial and ethnic groups are needed. Perhaps, large health care delivery systems and growing health system‐based networks are best suited to provide such results if their expanding electronic medical record and other data systems capture valid racial and ethnic designations and relevant clinical outcomes of their memberships.

Efforts to establish community surveillance for CHD harken back decades.102–103 A national system to monitor CHD incidence has never been established, however, and this gap has not gone unnoticed.104–107 As part of its recommendations, the Institute of Medicine highlighted the critical importance of having data on the incidence of CVD and the need for a system that would collect such data. The report cited potential avenues such as the establishment of registries, the use of cohort studies, and the use of claims and electronic medical record data to accomplish such a goal. The development of a national system to monitor the trend in the incidence of CHD would help to fill this current void in the knowledge base of the epidemiology of CHD and provide critical data to improve cardiovascular health of the US population.

In conclusion, definitive data about national trends of incident CHD in the United States currently are not available, and, therefore, clues about these trends must be gleaned from a variety of auxiliary data sources. Studies in different parts of the country demonstrate improvements in the incidence of CHD that may have commenced several decades ago in some parts of the country, and an increasing number of recent studies have described favorable trends during the first decade of the 21st century. Taken together, these studies yield encouraging but tentative signals that the incidence of CHD in the United States may be waning. Bringing greater clarity to this important topic of cardiovascular epidemiology poses a pressing public health need.

Show MeSH
Related in: MedlinePlus