Accuracy of Emergency Department Chest Pain Patients' Reporting of Coronary Disease History.

Introduction: History is an important component of emergency department risk stratification for chest pain patients. We hypothesized that a significant portion of patients would not be able to accurately report their history of coronary artery disease (CAD) and diagnostic testing.
Methods: We prospectively enrolled a convenience sample of a cohort of adult ED patients with a chief complaint of chest pain. They completed a structured survey that included questions regarding prior testing for CAD and cardiac history. Study authors performed a structured chart review within the electronic medical record for our 6-hospital system. Results of testing for CAD, cardiac interventions, and chart diagnoses of CAD/acute myocardial infarction (AMI) were recorded. Categorical data were analyzed by Chi-square and continuous data by logistic regression.
Results: About 196 patients were enrolled; mean age 57 ± 15 years, 48% female, 67% Hispanic, 50% income <$20,000/year. About 43% (95% confidence interval [CI] 35%-51%) of patients stated that they did not have CAD, yet medical records indicated that they were CAD+. With increasing age, patients were more likely to accurately report the absence of CAD (P < 0.001). There was no association between patients reporting no CAD, but CAD+ in records with respect to the following characteristics: female gender (P = 0.37), Hispanic race (P = 0.73), income (P = 0.41), less than or equal to high school education (P = 0.11), and private insurance (P = 0.71). For patients with prior AMI, 7.2% (95% CI 2.7%-11%) reported no prior history of AMI. Conclusions: Within our study group from a predominantly poor, Hispanic population, patients had a poor recall for the presence of CAD in their medical history. Copyright:

INTRODUCTION

Chest pain is the second most common overall chief complaint in the US Emergency Departments (EDs) after abdominal pain, comprising 5.2% of all visits recorded in the NHAMC database in 2011 and the most common among men over 65.[1] Emergency physicians must be able to effectively evaluate and risk stratify patients for severe cardiopulmonary disease, chief among them coronary artery disease (CAD) and related complications. Although emergency physicians receive extensive training in chest pain triage, they inadvertently discharge up to 2.1% of patients with AMI from the ED.[2]

Physicians utilize the past medical history and patients’ knowledge of their risk factors for CAD in their evaluation of a patient presenting to the ED with chest pain to determine safe triage. Researchers have shown that the past diagnosis of CAD and abnormal stress test is one of the most specific findings on a history for a diagnosis of acute coronary syndrome (ACS), even more specific than electrocardiogram (ECG) changes.[3] Investigators have derived well-validated clinical decision rules such as the HEART Score, HEART Pathway, and Emergency Department Assessment of Chest Pain Score (EDACS) for this risk stratification, and patient reported history is a key component of these tools.[34]

Emergency physicians may also use patient's previous visits and outpatient evaluation as a part of a comprehensive chest pain evaluation and risk stratification strategy. Cardiologists and inpatient medicine physicians perform cardiac catheterizations and cardiac stress tests, whether performed with ECG and a treadmill or, as is increasingly more common, myocardial perfusion scan to provide an objective assessment of patients’ disease burden (or in the case of cardiac catheterization, for therapy).[5678] However, it is unclear if patients are able to accurately report the results of such tests to the emergency physician.

Researchers have described patients’ ability to recall their diagnoses and risk factors for several other diseases such as CHF and COPD. Investigators have also examined the general knowledge of CAD and risk factors among both healthy and inpatient populations, but this general knowledge is of little utility to the bedside emergency physician evaluating for ACS.[910111213] We performed a prospective cross-sectional study to test the hypothesis that a significant proportion of ED patients reporting a negative history of CAD, ACS, and related testing would have evidence in the medical record that contradicts their recall.

METHODS

Study design

We prospectively enrolled a cohort of patients with chest pain to evaluate the accuracy of their recall for CAD/ACS diagnoses and diagnostic testing in ED patients.

Setting

We conducted our study in the Emergency Departments of CHRISTUS Spohn Memorial Hospital and CHRISTUS Spohn Shoreline Hospital in Corpus Christi, Texas. These facilities are the major teaching affiliates of Texas A and M Health Science Center and serve an inner-city population. The annual ED census of Memorial was 36000 and that of Shoreline was 45,000. Our institution's EMR includes >70% of all hospital-based ED visits within a 12-county region including 6-intrasystem facilities. Our study was approved by the CHRISTUS Health Institutional Review Board before the initiation of data collection.

Population

Our study included a convenience sample of medically stable, consenting, adult patients age >18 years that presented to the ED with a chief complaint of chest pain. Patients were excluded for any of the following reasons: refusal to provide written consent, pregnancy, incarcerated, inability to complete the questionnaire due to clinical instability, severe pain, or disorientation as determined by a study physician. Patients who were initially unstable or in severe pain were eligible for inclusion once their condition stabilized.

Study protocol

For a period of 6 months, patients were consecutively enrolled at hours, in which trained research associates (college students) were available to assist with data collection as well as by resident and attending physicians on duty in the ED at both locations. These hours were variable and included overnight as well as weekend days. Enrolled patients completed a written survey providing demographic information including sex, age, race, income, and education as well as answers to questions about their personal history of CAD, ACS, diagnostic testing, and interventions. Study authors then performed a structure chart review in our EMR for these same diagnoses, tests results, and interventions for each study participant. Our EMR covers a 6-hospital system which accounts for over 70% of ED visits in a 12-county area.

Statistical methods

Categorical data are presented as the frequency of occurrence and analyzed by Chi-square. Continuous data were presented as ± standard deviation and analyzed by t-tests. About 95% confidence intervals (CIs) and odds ratios were provided. Alpha was set at 0.05. Our primary outcome was the percentage of patients reporting no history of CAD who had evidence of CAD in our records, whether by testing or history. For the purposes of our study, a “positive history” of CAD was defined as the patient reporting or EMR evidence of one of the following: diagnosis of CAD or ACS, positive stress test, or positive angiography. We chose to include those with positive stress test and negative confirmatory angiography as confirmatory angiography was not universally available, and an ED physician presented only with a positive stress test result would presume a diagnosis of CAD. To assess interrater reliability, a second blinded reviewer independently collected key outcome data points for a sample of fifty patient records.

RESULTS

We assessed 217 patients for the possible enrollment with 21, respectively, either meeting exclusion criteria or refusing to participate. Thus, the final study group comprised 196 study participants. Table 1 provides a detailed summary of patient characteristics within that group. While our population was well balanced with respect to gender, patients were predominantly Hispanic, from lower-income groups and of lower educational status.

Table 1

Study population characteristics

Characteristic	Number (%)
Total patients	196 patients included
Mean age (years)	56.6±15.04
Gender (female)	48.21
Race
Hispanic	131 (67.18)
Nonwhite	145 (74.36)
Income
<$20,000 annual income	97 (50.52)
Education
High school education or less	136 (69.38)
Insurance
Medicare/medicaid	64 (33.68)
Uninsured	23 (11.92)

We found that 43% (95% CI 35%–51%) of patients who denied a history of CAD on our survey had evidence of coronary disease in the EMR. Table 2 summarizes the characteristics of patients stratified by the recall for a diagnosis of coronary disease and their medical record findings. There was no difference found between sexes (P = 0.37), Hispanic versus non-Hispanic race (P = 0.73), income level (P = 0.41), less than high school versus high school or more education (P = 0.11), or private insurance versus uninsured/government insurance (P = 0.71). We did find that older patients were more likely to accurately report the presence of CAD in their history, with an OR of 0.044 (P < 0.001). Investigators calculated this OR across the entire age span of patients in the study (20–89 years). Six of our patients had a negative catheterization in our record but were listed as positive for CAD. They are discussed below.

Table 2

Patients reporting no history of coronary disease

	Patients answering “no” to CAD	No evidence of CAD in EMR (%)	Evidence of CAD in EMR (%)
Overall, 95% CI	144	82 (56.94)	62 (43.06), 0.35-0.51
Sex (P=0.37)
Male	73	44 (60.27)	29 (39.73)
Female	70	37 (52.86)	33 (47.14)
Race (P=0.73)
Hispanic	101	57 (56.44)	44 (43.56)
Nonhispanic	42	25 (59.52)	17 (40.48)
Income (P=0.412)
<$20 k	74	40 (54.05)	34 (45.95)
$20–40 k	23	10 (43.48)	13 (56.52)
$40–60 k	15	11 (73.33)	4 (26.67)
$60–80 k	12	7 (58.33)	5 (41.67)
>$80 k	17	11 (64.71)	6 (35.29)
Education (P=0.11)
Less than HS	49	23 (46.94)	26 (53.06)
HS or more	68	42 (61.76)	26 (38.24)
Insurance type (P=0.71)
Public/uninsured	83	46 (55.42)	37 (44.58)
Private	58	34 (58.62)	24 (41.38)

CI: Confidence interval, CAD: Coronary artery disease, EMR: Electronic medical record, HS: High school

Table 3 summarizes the characteristics of patients stratified by the recall for a prior diagnosis of AMI and their medical record findings. Among patients who denied having an AMI in their history, 7.2% (95%CI 2.7%–12%) had a diagnosis of AMI in the EMR. There was no significant difference found with regard to sex, age, Hispanic versus non-Hispanic race, income level, education level, or insurance type [Table 3]. In terms of diagnostic testing, of those who denied having a stress test [Table 4], 13.2% (95%CI 5.2%–21.3%) of study participants had evidence of a stress test in our records, whether performed in our facility or referred to in cardiology notes. A significantly higher percentage of Hispanics, 18%, had evidence of stress testing when they reported a negative history when compared to non-Hispanics (P = 0.038). For those study, patients denying a cardiac catheterization in the past [Table 5], 8.49% (95%CI 3.2%–13.8%) had evidence in our records. No significant difference was found with sex, age, race, income level, education level, or insurance type.

Table 3

Patients reporting no history of myocardial infarction

	Patients answering “no” to MI	No evidence of MI in EMR (%)	Evidence of MI in EMR (%)
Overall, 95% CI	125	116 (92.8)	9 (7.2) 0.025-0.153
Sex (P=0.26)
Male	60	54 (90.0)	6 (10.0)
Female	64	61 (95.31)	3 (4.69)
Race (P=0.062)
Hispanic	92	83 (90.22)	9 (9.78)
Nonhispanic	30	30 (100.00)	0
Income (P=0.336)
<$20 k	59	56 (94.92)	3 (5.08)
>$20	62	56 (30.32)	6 (9.68)
Education (P=0.933)
Less than HS	43	40 (93.02)	3 (6.98)
HS or more	61	57 (93.44)	4 (6.56)
Insurance type (P=0.91)
Public/uninsured	70	65 (92.86)	5 (7.14)
Private	52	48 (92.31)	4 (7.69)

MI: Myocardial infarction, CI: Confidence interval, EMR: Electronic medical record, HS: High school

Table 4

Patients reporting no history of stress test

	Patients answering “no” to stress test	No evidence of stress test in EMR	Evidence of stress test in EMR
Overall, 95 CI	68	59 (86.76)	9 (13.24) 0.052-0.213
Sex (P=0.76)
Male	41	36 (87.8)	5 (12.2)
Female	27	23 (85.19)	4 (14.81)
Race (P=0.03)
Hispanic	48	39 (81.25)	9 (18.75)
Nonhispanic	20	20 (100.00)	0
Income (P=0.336)
<$20 k	39	32 (82.05)	7 (17.95)
>$20	27	26 (96.30)	1 (3.70)
Education (P=0.933)
Less than HS	30	24 (80.00)	6 (20.00)
HS or more	24	21 (87.50)	3 (12.50)
Insurance type (P=0.91)
Public/uninsured	47	39 (82.98)	8 (17.02)
Private	18	17 (94.44)	1 (5.56)

CI: Confidence interval, EMR: Electronic medical record, HS: High school

Table 5

Patients reporting no history of cardiac catheterization

	Patients answering “no” to Cath	No evidence of Cath in EMR	Evidence of Cath in EMR
Overall, 95 CI	106	97 (91.51)	9 (8.49), 0.032-0.138
Sex (P=0.661)
Male	54	50 (92.59)	4 (7.41)
Female	51	46 (90.20)	5 (9.80)
Race (P=0.898)
Hispanic	72	66 (91.67)	6 (8.33)
Nonhispanic	33	30 (90.91)	3 (9.09)
Income (P=0.06)
<$20 k	54	52 (96.30)	2 (3.70)
>$20	49	42 (85.71)	7 (14.29)
Education (P=0.707)
Less than HS	35	32 (91.43)	3 (8.57)
HS or more	47	44 (93.62)	3 (6.38)
Insurance type (P=0.67)
Public/uninsured	64	59 (92.19)	5 (7.81)
Private	39	35 (89.74)	4 (10.26)

CI: Confidence interval, EMR: Electronic medical record, HS: High school, Cath: Catheterization

A second blinded reviewer abstracted data for key end points for a sample of fifty patients. We subsequently found the following kappa values: evidence of CAD in the record (0.53, moderate agreement), evidence of AMI in the record (0.71, substantial agreement), evidence of a stress test result in the record (0.68, substantial agreement), evidence of a coronary catheterization in the record (0.51, moderate agreement), evidence of percutaneous coronary intervention/coronary artery bypass grafting (PCI) in the record (0.76, substantial agreement).

We included six patients in the primary end point analysis with a negative catheterization. The first patient had a stress test and catheterization done within our system, but at a later date was diagnosed with two MIs and had one vessel PCI performed at an outside facility. The second was diagnosed with CAD in the chart with the following negative catheterization. However, the patient carries a diagnosis of Lupus and was diagnosed with an MI in 2016, before the catheterization. Patient 3 had a negative catheterization and stress test but carried diagnoses of myocardial infarction (MI) and angina. Patient 4 had a positive stress test and negative catheterization at our facility; however, cardiology notes referred to an outpatient catheterization with >70% stenosis. Patients 5 and 6 had negative cardiac catheterizations at our facility but were diagnosed with MI at a later date. All of these patients are at increased risk for ACS in the acute setting and for the purposes of ED risk stratification are understood to effectively have CAD, whether by listed diagnosis of ACS, reference to outpatient positive cath, or other high-risk condition for vascular pathology systemic lupus erythematosus.

DISCUSSION

We found that a significant proportion of patients in our study group were unable to accurately recall their medical history as it relates to the presence/absence of coronary disease and diagnostic testing for such. Our findings are congruous and reflect, perhaps, a logical extension of a general lack of awareness by patients of coronary risk factors and objective tests utilized for cardiac risk stratification. Almas et al. surveyed 200 nonmedical university students and found that whereas most were aware that “heart disease” was a leading cause of death and smoking a major risk factor, less than half could correctly define a cardiac catheterization.[11] Further, Kayaniyil et al. surveyed 1308 cardiac inpatients in Ontario, and they found that overall knowledge of signs/symptoms and risk factors for CAD was poor, particularly for those patients with lower household income and educational attainment.[12]

Physicians depend on a reliable history for the risk stratification for both ACS and non-ACS-related cardiopulmonary diseases. Allen et al. conducted a prospective survey of 122 ambulatory CHF patients and compared their estimation of life expectancy to a model predicted estimate.[9] They found that patient expectations and the model were highly discordant with mean patient estimates more than 3 years higher. Similarly, Barr et al. surveyed a national sample of 1003 COPD patients regarding their disease. The authors found that only 10% of surveyed patients knew their own forced expiratory volume, which is a known predictor of morbidity and mortality in this disease process.[10] Allen et al. raised the concern that patients with poor knowledge and insight into their own illness are at risk to choose costly advanced therapies in the face of poor prognosis.

In a somewhat analogous manner, our results suggest that an emergency physician facing a patient with chest pain that does not have medical records within electronic access could be relying on inaccurate recall/poor comprehension that leads to an erroneous risk assessment by the physician for ACS and adversely impact the potentially costly decision to admit or discharge the patient. Fanaroff et al. in a systematic review reported that prior positive angiography/stress test and prior MI were independent predictors of ACS in patients presenting for an initial evaluation of ACS-related symptoms.[3] An additional risk factor can in many cases raise a HEART Score from 3 to 4, increasing the risk for major adverse cardiovascular events at 6 weeks from 1.7% to a minimum of 12% based only on history, exam, ECG, and initial troponin.[1415]

We believe that our study is novel for our intent to evaluate patients’ recall in the ED setting for prior diagnosis and/or testing for ACS/CAD. This represents a true-to-life reproduction of a situation where a physician or health-care provider must elicit an accurate history and may not have the benefit of an inclusive EMR. We find it especially concerning that 8.5% of patients in our study did not remember and/or understand that they had undergone cardiac catheterization, a relatively invasive procedure.

Interestingly, we found that patients who reported a prior diagnosis of AMI did not necessarily understand that they have CAD. Within our study group, 144 patients denied a diagnosis of CAD, whereas 125 denied prior AMI. About 19 patients then reported a positive history of MI, but either did not know or understand that AMI is related to underlying disease of the coronary arteries. We believe this represents an opportunity for patient education for cardiac care providers.

Limitations and future questions

Our study has several limitations. As we were not able to access the EMR from other outpatient and inpatient health facilities in the area, we suspect that the reported proportion of patients inaccurately denying their history of disease and testing may be somewhat underestimated. We believe almost any similar study would suffer from the same limitation. However, we benefited in our investigative efforts due to the large percentage of overall ED/hospital visits that are served by our hospital system within a 12-country region.

We also recognize that our study results may not be generalizable to other settings. Our population is predominantly lower income, Hispanic, and has attained a relatively low educational level. We believe it will be important for future investigators to examine the question of patient recall/CAD disease and testing accuracy within more diverse populations.

We believe our results suggest several avenues for future investigation in this area, especially with respect to patient education. It is important for patients to understand their disease processes both for ethical reasons and to help them make educated decisions about treatment and utilization of the ED. Furthermore, our data provide a starting point for conversations at a local and larger level with our primary care and cardiology colleagues. We need to discuss interventions in the ED, clinic, and hospital to improve patient recall and go on to evaluate their effectiveness.

CONCLUSIONS

Within our study group from a predominantly poor, Hispanic population, patients had a poor recall for the presence of CAD in their medical history. Future investigation is warranted to explore pathways to improve patient recall and comprehension of their cardiac testing and disease history.

Research quality and ethics statement

This study was approved by the Institutional Review Board/Ethics Committee CHRISTUS Health System study #2016-037. The authors followed applicable EQUATOR Network (”http://www.equator-network.org/) guidelines during the conduct of this research project.ing the conduct of this research project.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.