Introduction

Coronary artery disease (CAD) is the leading cause of death of Americans with an estimated 650,000 a year experiencing a myocardial infarction (MI)₁. Because of the high prevalence of the disease, public awareness has resulted in an increasing number of people who present to the emergency department for evaluation of chest pain. The medicine wards at the study site, a public hospital, have noted a rise in patient census, due in part to this increase. It is estimated that $13 billion is spent annually in the US on hospitalizing patients with chest pain who turn out not to have a MI ₂. The current treatment practice at the study site involves admission for provocative testing (Stress Testing with Nuclear Imaging most often) after an acute coronary syndrome (ACS) is ruled out. However, these tests are only done during the “work week” causing delays during hospital admission. It is feasible, if these patients are evaluated with risk stratification that some of them could be discharged with outpatient testing. Therefore, patients would not have extended hospitalizations waiting on a stress test. The benefits from this risk stratification are multiple, including time involved, resources utilized, and money spent.

It is the standard of care in the United States to utilize risk factor stratification to help determine the type of cardiac work-up a patient needs when presenting with chest pain. In many facilities a scale is used to determine the risk of an acute coronary syndrome based on risk factors. One example is the Thrombolysis in Myocardial Infarction (TIMI) risk score. The TIMI score is based on risk factors for CAD and clinical presentation during a possible ACS₃.

For many years it was felt that approximately 50% of coronary events occurred in patients with no risk factors. However, in reality, the presence of one or more risk factors in the young (18-59) is highly sensitive for future coronary events₄. The traditional risk factors for CAD are advanced age, sex, family history, dyslipidemia, cigarette smoking, hypertension, diabetes, obesity, and sedentary lifestyle. Multiple studies have been conducted to show potentiation effects of risk factors on a person's risk for having CAD.

The TIMI score was developed by the TIMI trials that began in 1984. They examined thrombolytic and antithrombotic regimens in acute MI and unstable angina (UA). Currently there have been 23 TIMI trials that have provided insights into the pathophysiology, and clinical course and have provided information that is useful in the treatment of acute MI and UA₅. In the TIMI-11B trial, a score was developed based on risk factors and clinical presentation that distinguished those patients with UA or non-ST elevation myocardial infarction (NSTEMI) into risk categories of low (0-2), intermediate (3-4) and high (5-7). These risks were associated with the chance of having an acute coronary event in the following 14 days after presenting with chest pain₃.

A retrospective study was published in April 2004 that looked at the correlation of the TIMI risk score with the extent of native vessel disease in patients diagnosed with NSTEMI and who underwent angiography. The results indicated that the TIMI score did correlate with the extent of coronary disease confirmed by coronary angiography and 57% of the low risk patients had nonsignificant CAD or normal angiogram. This study showed that the TIMI risk score has a broader usefulness and can be used at the bedside when deciding on diagnostic and therapeutic interventions₆.

The American College of Cardiology (ACC) and the AHA have published together in their practice guidelines, an algorithm for the “Evaluation and Management of Patients Suspected of Having an ACS” ₇. If the cardiac enzymes remain negative and the patient has no further episodes of chest pain, then a stress test is recommended. Stress test is defined as an Exercise Treadmill Testing (ETT), with or without imaging modality, or Dobutamine/Adenosine Myocardial Perfusion Imaging (MPI). The guidelines indicate that admission is not necessary for those at low risk based on risk factor stratification and stress testing may be done prior to discharge or as an outpatient.

However, there is some concern that risk stratification may not be safe for those people who are in lower socioeconomic circles, as defined by lower income and educational levels. There has been documentation over the past ten years that seems to indicate there is an inverse relationship between SES and cardiovascular risk₈. All risk factors do not have the same association. However, hypertension, smoking, BMI, and physical inactivity do seem to correlate. Of note, elevated cholesterol and lipids are not associated₉.

Environmental factors also play a role in increasing the risk for CAD. When comparisons were made between individuals from New York and their areas of birth, the south was shown to have higher incidence of CAD, which supports the impact of early life influences₉.

The purpose of this study is to determine if the A-TIMI scoring system can differentiate a low risk patient from an intermediate risk patient for safe discharge with outpatient provocative testing at a public hospital. Specifically, an inner city hospital located in the south, that mainly provides medical care to a lower socioeconomic class known to have more adverse cardiac outcomes.

Methods

Participants

Procedure

Instrument

Data Analyses

Results

This study was conducted over a twelve-month period where 145 participants met the entry parameters. The demographic findings are reported in Table 1 and the study results are noted in Table 2.

Figure 1

Table 1: Demographic Findings

Figure 2

Table 2: Study Findings

Females comprised 78 (or 53.79%) and males comprised 67 (or 46.21%) of this population. Average overall age was 49, with African-Americans, Caucasians, and Hispanics composing 45.52%, 53.79%, and 0.69%, respectively. Participants spent an average of 2.3 hospital days with 36 (24.83%) of the hospital admissions greater than or equal to four days due to complications or delays in procedures. Ninety-one percent of this population was calculated as low-risk patients, and of these 132 participants, 77 had negative stress tests and 55 had positive stress tests. Thirteen were calculated as intermediate-risk with only three positive stress tests vs. ten negative. Of the 58 total positive stress tests (both low and intermediate risk), only 26 (17.93%) were taken for left heart catheterization (LHC) as deemed by the Cardiology Service at the study site or because of refusal of procedure (2). Eleven of the 26 LHCs did not have findings consistent with the stress tests, therefore the stress tests were termed false positives. The remaining 15 had findings as indicated on the stress tests, and were termed true positives. Only two patients were referred for coronary bypass surgery. And interventions, via percutaneous angioplasty or stent placements, were performed on 6 of the 15. Thirteen received medical management or ischemia was so minimal that Cardiology felt the chest pain experienced by the patient was not cardiac in origin. Eight had fixed defects and no evidence of ischemia, but indicating previous infarct. Statistics of this population regarding positive risk factors are seen in Table 2 and 3.

Figure 3

Table 3: Risk Factor Prevalence in Study Population

Only 28 (19.31%) of this study population had funding for medical care as provided by Medicaid, Medicare, or private insurance.

None of the above findings were statistically significant [see Graph 1, 2, & 3].

Figure 4

Graph 1: Comparison of positive and negative stress tests, between low and intermediate risk categories.

Figure 5

Graph 2: Comparison of positive and negative stress tests, between low and intermediate risk categories in the female study population.

Figure 6

Graph 3: Comparison of positive and negative stress tests, between low and intermediate risk categories in the male study population

The only significant finding was noted when comparing males to females [see Graph 4]. It was noted that 19 of the 58 positive stress tests were women (33%), while 39 were men (67%). When comparing positive and negative stress tests against gender, there was a statistical significance with a P value of less than or equal to .001.

Figure 7

Graph 4: Comparison of positive and negative stress tests, between males and females.

Discussion

At this study site that serves the lower socioeconomic population, a fairly equal number of men and women presented to the emergency department or clinic for evaluation of chest pain. There was no correlation between intermediate or low risk patients and whether or not they had a positive stress test, but when evaluated as a whole, there was a correlation, however, it was statistically insignificant.

A notable finding was the difference between men and women and the chance of having a positive stress test. One in four women (24.36%) were likely to have a positive stress test, indicating that risk stratification may be acceptable in the female segment of this population. However, the men had a one in 1.72 chance of having a positive stress test (58.21%), and therefore, risk stratification may not be acceptable in the male segment of this population

Problems

There was only one finding that was of statistical significance in this study, due in part to the sample size. Bias also played a big part in this study.

First, the interpretation of the nuclear images from the stress tests was dependent on one individual, the nuclear radiologist. Two individuals, at six months intervals, were responsible for reading the stress test results during the course of this study.

Second, the Cardiology staff physician determines if a LHC is necessary when a positive stress test occurs. Four of the 13 classified as “medical management”, were actually the cardiologist's opinion that chest pain was not from CAD despite a positive stress test, therefore a LHC was not performed. Since all positive stress tests were not taken for LHC, the false positive stress tests cannot be included in the “negative” tests, possibly effecting percentages and thus statistical significance.

Other problems noted concerned information gathering from the participant. Risk factor scoring was determined by taking a history from the participant. Some denied a history, such as dyslipidemia, when a previous lipid panel was documented in the computer. The family history of CAD posed a problem because many people would associate chest pain as having a MI. Ages regarding CAD history were not consistently documented by the admitting physician, therefore making it hard to determine if a primary relative actually had a history of CAD (i.e. MI <55).

Acknowledgements

Footnotes