Myocardial infarction classification and its implications on measures of cardiovascular outcomes, quality, and racial/ethnic disparities

Abstract Heart disease continues to be the leading cause of death in the United States, with approximately 805 000 cumulative deaths from myocardial infarctions (MI) from 2005 to 2014. Gender and racial/ethnic disparities in MI diagnoses are becoming more evident in quality review audits. Although recent changes in diagnostic codes provided an improved framework, clinically distinguishing types of MI remains a challenge. MI misdiagnoses and health disparities contribute to adverse outcomes in cardiac medicine. We conducted a literature review of relevant biomedical sources related to the classification of MI and disparities in cardiovascular care and outcomes. From the studies analyzed, African Americans and women have higher rates of mortality from MI, are more probably to be younger and present with other comorbidities and are less probably to receive novel therapies with respect to type of MI. As high‐sensitivity troponin assays are adopted in the United States, implementation should account for how race and sex differences have been demonstrated in the reference range and diagnostic threshold of the newer assays. More research is needed to assess how the complexity of health disparities contributes to adverse cardiovascular outcomes. Creating dedicated medical quality teams (physicians, nurses, clinical documentation improvement specialists, and medical coders) and incorporating a plan‐do‐check‐adjust quality improvement model are strategies that could potentially help better define and diagnose MI, reduce financial burdens due to MI misdiagnoses, reduce cardiovascular‐related health disparities, and ultimately improve and save lives.


| INTRODUCTION
In the United States, heart disease continues to be the leading cause of death in men and women regardless of race or ethnic group. [1][2][3] With respect to racial/ethnic groups in the United States, collectively African American adults experience a higher rate of cardiovascular diseases, compared to white adults. 4 This is expected to continue into the future, as it is estimated that 45.1% of the US population will have some form of heart disease by 2035. 5 Unless the prevalence of heart disease is better controlled and significantly reduced, total direct medical care costs related to heart disease are expected to increase from $318 billion to $749 billion. 5 Within the spectrum of heart disease, someone in the United States has a myocardial infarction (MI) every 40 seconds. 5 There have been recent discussions involving the misdiagnosis of MI and its implications on patient care and medical costs.
The reclassification of the International Statistical Classification of Diseases and Related Health Problems (ICD) 10 codes in 2017 helped further define the types of MI, yet more work is needed to educate medical professionals on a more useful way of making a bedside clinical diagnosis of MI and prevent medical coding errors. Further, elevated cardiac troponin (cTn) has been associated with disparities in race and gender-related care. 2,6 Inconsistencies in clearly distinguishing a type of MI, widening racial and ethnic gaps in health, and gender-related differences in biomarker analyses contribute to the incongruency of diagnosis and medical coding of MI, leading to unnecessary costs and death. Currently there is not enough data on the nexus of MI misdiagnoses and health disparities, and how both attribute to medical coding and cardiac medicine as a whole. Thus, it is imperative to get a better understanding of these challenges from a quality perspective to improve patient treatments and outcomes. We will highlight some of these challenges and present potential strategies to address these challenges in this review.

| DISTINGUISHING DIFFERENCES IN MI CLASSIFICATION
Despite efforts to improve health among disadvantaged racial and ethnic groups in the United States, health disparities continue to exist.
The interactions between the many factors that attribute to these disparities, and how they result in different outcomes, are not always well defined. Thus, before discussing health disparities of racial/ethnic groups with respect to MI one must first dissect the disparities within the classification of MI. The standard clinical definition of MI signifies the presence of an acute myocardial injury detected by abnormal cardiac biomarkers, namely cTn, in the setting of evidence of acute myocardial ischemia. 7 A precise, definitive diagnosis not only provides valuable guidance on treatment, patient prognosis, and lifestyle changes, but it is also associated with policy and resource allocation, medical coding-related diagnosis, and hospital reimbursement. 7,8 Due to the secondary characteristics of multiple cardiac events, it is necessary to further delineate and define the various types of MI for proper diagnosis. Type 1 MI is an MI with clinical evidence of ischemia caused by atherosclerotic plaque disruption resulting in coronary thrombosis and detection of a rise and/or fall of cTn values with at least one value above the 99th percentile upper-reference limit (URL). 7,8 In addition, patients must also exhibit one of the following symptoms of myocardial ischemia conditions: new ischemic electrocardiogram changes, development of pathological Q waves, evidence of new loss of viable myocardium or new regional wall motion abnormality in a pattern consistent with ischemic etiology via imaging, and identification of a coronary thrombus by angiography or autopsy. 7 Type 2 MI also includes the aforementioned symptoms, but is due to a supply-demand mismatch of myocardial oxygen in the absence of coronary thrombosis. 7,8 Prior to 2017, differentiating patients with type 1 and type 2 MI-based via medical coding was not possible because an ICD code for each particular MI subtype did not exist. 8 Without strict ICD 10 coding criteria and an effective means of determining the specific type of MI at bedside and given that patients with type 2 MI can potentially have numerous underlying comorbidities, a type 2 diagnosis is subject to uncertainty or misdiagnosis. An estimated 90% of type 2 MI patients were not being coded. 8,9 An ICD code now exists for type 2 MI, yet issues still surround accurately documenting other forms of type 1 MI as type 2 MI. Another potential misclassification exists between myocardial injury and MI.

| Myocardial injury vs MI
Myocardial injury is defined as elevated cTn values, at least one value above the 99th percentile URL, and is considered acute if there is a rise and/or fall of cTn values. 7 MI is a subset within acute myocardial injury.
In this review, the use of "injury" throughout the manuscript is intended as injury without MI. The most current universal definition of MI states the importance of the differentiation of MI from myocardial injury and further differentiation from other types of MI, especially type 2 MI. 7 The definition of acute myocardial injury states that it is characterized by a rise and/or fall of cTn, the most agreed upon standard being >20% change for patients with normal baseline cTn or a > 50% change for patients with a baseline elevated Tn; however there is currently still discrepancy on exactly how much cTn is needed in order to distinguish the injury as acute. A number of patients with elevated cTn have no clinical evidence of ischemia (ie, no ischemic chest symptoms and no ECG changes). Without ischemia, a diagnosis of type 2 MI should not be made. 9 Yet, in a single-center observational study, McCarthy et al found that approximately 42% of type 2 MI patients were misdiagnosed and actually had myocardial injury without ischemia. 10 Currently, there is no ICD 10 code for myocardial Injury. A specific ICD 10 code for myocardial injury would reduce misdiagnosis and improve the characterization of a patients' condition. Myocardial injury is also a recent focus of many quality improvement and value-based programs.
Recent emphasis on quality improvement efforts adhering to guidelines and ensuring proper MI diagnosis have been connected to performance metrics and reimbursement. One such effort, the Hospital Readmission Program (HRRP), was designed to improve the quality of patient post-acute care and reduce Medicare spending by preventing rehospitalizations of conditions such as acute MI. 8 Acute MI is one of six conditions included in the HRRP, with a combined total of $528 million being withheld from hospitals due to readmissions in 2017. 8 Of all readmitted patients with MI contributing to the HRRP penalty, about 10% may have actually had type 2 MI. 8,11 In conjunction with the Hospital Value-Based Purchasing Program, and after strict adjudication with physician medical record reviewers using the most current universal definition of MI, 7 McCarthy et al found that approximately 41% of patients with nonischemic myocardial injury are frequently misdiagnosed and subsequently billed as having type 2 MI. 10 Compared to patients with myocardial injury, patients type 2 MI also exhibited a higher prevalence of cardiovascular-related comorbidities such as coronary artery disease (50.6% vs 33.2%; P < .001), heart failure (52.4% vs 37.4%; P < .001), peripheral arterial disease (23% vs 12.1%; P < .001) and prior MI(21.8%vs 14.3%; P = .02). 10

| STEMI, NSTEMI, and type 2 MI
Type 1 MI can be further divided into two classifications: ST-elevation myocardial infarction (STEMI) and non-ST-elevation myocardial infarction (NSTEMI). STEMI is defined as an acute coronary thrombosis or persistent ST-segment elevation of ≥1 mm in ≥2 contiguous electrocardiographic leads. 7,12 NSTEMI is defined as ischemic symptoms at rest from an acute coronary plaque rupture or erosion, lasting ≥10 minutes, occurring within 24 hours before hospital admission, and displaying either elevated cardiac biomarkers (either creatine kinase or cTn) within 24 hours after initial presentation. 12 Prior to the introduction of the ICD-10 code for type 2 MI, patients with type 2 MI were often coded as a NSTEMI or with no code at all. 9,13 While STEMI care is well defined and fairly consistent in delivery, NSTEMI management is substantially more variable on the patient, clinician, facility, and regional levels.. 4 Since there is now a specific code for type 2 MI, and type 2 MI does not feature acute coronary thrombotic plaque disruption, great efforts should be made by clinicians to not identify it as NSTEMI. 13 Doing such could have adverse effects on patient prognosis, treatment, and outcome. Improved methods of distinguishing type 1 and 2 MI could further enhance the study of MI and the potential development of MI-specific treatments.

| DISPARITIES IN MI OUTCOMES AND CLASSIFICATION
Tackling health disparities can be daunting given that the work involves not only patient-doctor interactions, but also how this interaction involves patient personal preferences, trust level, education, and socioeconomic factors; unfortunately, these interactions are often not well characterized. 3 In addition, many physicians may have little to no training on how to incorporate or even recognize health disparities. Some physicians may also shift the sole responsibility of addressing health disparities to the society as a whole or the government. Despite the ongoing efforts of government entities such as the Institute of Medicine and the National Institute of Minority Health, African Americans still carry the highest burden of heart disease and mortality rates compared to white Americans. 3,4,14 African Americans also have higher mortality rates from MI. 15  Both studies highlight not only MI disparities in African Americans collectively, but also gender differences. Heart disease is the number one killer of women worldwide regardless of race or class. 2,16 Both biological and social factors contribute to the gender disparities seen in heart disease, as there are differences in the pathobiology of acute MI which lead to misdiagnosis. 2 This appears to also be the case for African Americans as a whole, especially as it pertains to the higher prevalence of comorbidities in this population which both increase the incidence of MI and the adverse effects after MI. 17 12 NSTEMI was found to be more prevalent than STEMI in African American patients. 12 African American patients were also younger yet had a higher incidence of comorbidities such as hypertension, diabetes mellitus, and stroke. 12 Interestingly gender differences were also evident, as African American women had a higher frequency of diabetes mellitus compared to their counterparts. 12 While acute medical treatment was similar among African American and white women, African American men had significantly lower overall rates of being prescribed an anticoagulant. 12 Table 1  , and there was a significant trend toward reduced timely reperfusion among patients in the lower quartiles than those in the higher quartiles. 23 Although the incidence of inhospital mortality among patients who had delayed or no reperfusion was higher than those who underwent timely reperfusion (P < .001), there was no significant difference in the inhospital mortality incidence across the SES quartiles. 23 However, patients in the lower quartiles who underwent timely reperfusion had a significantly higher incidence of inhospital mortality compared to those in the higher quartiles. 23 Women also were found to have had higher inhospital mortality rates compared to men across all SES quartiles (P < .001). 23 While this study found a higher proportion of nonwhite in the lower socioeconomic quartiles, it does not necessary delineate racial or ethnic disparities seen across various SES strata. However, it gives credence to the importance of including such data into medical charting to give context to a better understanding of patient health outcomes.

| Implications of cTn levels on disparities
The use of cardiac-specific biomarkers is not only critical for defining, providing specific diagnosis and/or prognosis, and ensuring accurate decision-making in the management of acute MI, but also provides information about associated pathways and/or consequences of MI, which may help to further understand population specific characteristics. 24,25 The discovery of cTn in 1963 created a paradigm shift in the ability to use a more sensitive cardiac-specific biomarker to optimize a diagnosis of MI. 7 However, using the reigning gold standard is not without caveats. Elevated cTn is not only specific for myocardial injury but also for the other various subtypes of MI and other comorbidities.
As mentioned previously, heart disease is the leading cause of death in women. Along with biological and social factors being key factors in  6 Thus, we recently hypothesized that patients with elevated cTn would undergo more testing than those without elevated cTn. Using a retrospective study cohort of 26 663 subjects, 18.6% had at least one elevated cTn assay while acute MI was diagnosed in 3.9%. 6 We found that men were more likely to undergo catheterization and cardiology consultation compared to women (OR 1.29, 95% CI 1.20 to 1.39 and OR 1.45, 95% CI 1.31 to 1.61). 6 In addition, African American patients were less likely to have either catheterization (OR 0.85, 95% CI 0.77 to 0.93) or consultation (OR 0.72, 95% CI 0.63 to 0.82) performed. 6 Although limited by our inability to derive a statistical method to determine the change is cTn (ie, a rise and/or fall) to accurately distinguish the type of MI, our data implies that an indiscriminate use of cTn may be incongruous with best clinical practice. It also stresses the continuous need to focus efforts on strategies to include social determinants of in the further development and adoption of high-sensitivity cTn assays.

| DISCUSSION
Accurate coding of MI is invaluable for ultimately improving patient prognosis and outcomes as well as assessing value-based programs and hospital quality metrics. 13 to give residents hands-on experiential training in QI. 28 Two key aspects of using this approach among these fellow were (1)  and CDI nurses to help identify key areas of shortcomings as needed.
By consistently using a PDCA cycle, the retention of improvement through standardization could increase over time, leading to a potential decrease of institutional financial burdens due to MI misdiagnosis, over or underuse of cardiovascular services and penalties of coding. Continuous QI practices could also increase awareness of health disparities and its impact on MI diagnosis and treatment; and, ultimately increase overall patient outcomes and health.

| CONCLUSION
The evolving and expanding epidemic of comorbidities such as obesity, hypertension, metabolic syndrome and diabetes disproportionately affects minorities and women, and these metabolic disturbances both increase the incidence of type 2 MI and the adverse outcomes after an MI. 12,17 Steps have been taken to explore the role of race and gender disparities in MI and biomarker thresholds on the proper identification and diagnosis of various MI subtypes and offer strategies to implement more efficient bedside feedback and documentation among cardiovascular medical teams. Creating dedicated interdisciplinary medical quality teams and incorporating a PDCA QI model are strategies that could potentially help reduce cardiovascular-related health disparities and ultimately improve and save lives.