A retrospective analysis of gender among patients admitted to a clinical decision unit at risk for acute coronary syndrome

INTRODUCTION

Chest pain is a common chief complaint and encompasses nearly 5.2% of all patients seen in emergency departments (ED). In the United States, this accounts for approximately six million annual visits.[1-4] The evaluation of patients with chest pain in the ED is a time-consuming and challenging process because the physician must identify those with life-threatening cardiac chest pain, while avoiding unnecessary and costly evaluations of those with non-cardiac chest pain. Emergency physicians (EPs) utilize risk stratification tools such as the HEART[5] score, which is an externally validated screening tool designed to identify ED patients with acute chest pain for early discharge. Utilization of the HEART score reduces the need for objective cardiac testing in the first 30 days, shortens the length of stay, increases early discharge,[5] and is part of the decision-making algorithm for patients admitted to the clinical decision units (CDU), also known as ED observation units. CDUs were established to help mitigate the burden of inpatient admission for low-risk patients who required an expedited cardiac work-up but did not necessarily warrant admission.[5,6] Many studies have shown the benefit of the CDU in mitigating the cost burden of patients presenting to the ED with low-risk chest pain and providing cardiologists with the ability to evaluate patients outside of the ED.[2, 6-8]

There are many studies regarding gender differences among patients presenting with acute coronary syndrome (ACS), high-risk chest pain and known coronary artery disease (CAD).[9-14] However, limited literature exists regarding the impact of gender among patients presenting with low-risk chest pain. It is clear that women with multiple risk factors who present with chest pain are at a higher risk of adverse outcomes, but it is unclear if the same rate of risk is present in women with low-risk chest pain.[15] As a result, many EPs have had a lower threshold to start an ACS work-up on women in the evaluation of chest pain and atypical chest pain.[16, 17] Previous studies found that in comparison to men, it was more common for women with ACS to present chest pain-free or with non-cardiac symptoms,[12,13] thus leading to a delay in identification and intervention. Patients with ACS who present without chest pain (regardless of gender) have higher morbidity and mortality.[14] One review found that 8.4% of 20,881 patients with ACS presented without chest pain, and these patients were more likely to be female, older or have comorbidities.[13] A recent study performed in Korea showed that in comparison to men, women present later in life with CAD and manifest more atypical symptoms such as palpitations and headaches.[11]

Few population-based studies have evaluated the difference in clinical outcomes of men and women with suspected ACS who were admitted to an observation unit for evaluation of low-risk chest pain.[4] The authors’ objective is to identify gender differences in disposition and provocative testing, defined as treadmill or nuclear stress testing, among patients admitted to the CDU for low-risk chest pain.

METHODS

Study design

This study was performed at a 100,000-visit, urban, teaching hospital ED. A retrospective analysis was performed on patients admitted with concerns of ACS to the CDU over the course of one year (2017). Patients were admitted to the CDU with complaints of chest pain or chest pain-related symptoms (n=732; supplementary Table1). These patients had been deemed low risk for an acute coronary event at admission to the CDU by the EP using clinical judgment. The protocol for the chart review was reviewed by the Institutional Review Board at Long Island Jewish Hospital, Northwell Health, and the protocol was deemed exempt from human subjects research. The HEART score was obtained from the electronic health record as calculated by the EPs at the time of admission to the observation unit. Outcomes measured included the rate of positive stress tests as identified by the interpreting physician in the final report, rate of admission from the CDU and rate of the requirement of the cardiac catheterization as judged by the cardiologist evaluating the patient prior to disposition from the CDU.

CDU description

The CDU consists of twelve beds and is staffed with 24-hour physician assistant (PA) and 12-hour physician coverage per day. The CDU approval was granted by the hospital’s Institutional Review Board. The authors performed a retrospective chart review of patients who were admitted to the CDU under the “Chest Pain with High Sensitivity Troponins” protocol (supplementary Figure 1). The authors evaluated the patient’s gender, HEART scores, and various outcomes including discharge or admission with cardiac catheterization.

Inclusion and exclusion criteria

The CDU inclusion criteria for the chest pain protocol included stable vital signs, as determined in the ED, troponin <0.06 ng/mL (4^th generation troponin I assay), anticipated length of stay > 4 h, and potential to be discharged in 23 h. Patients must be alert and oriented, able to ambulate or require minimal assistance, and carry no active psychiatric illness or dementia. Exclusion criteria for the chest pain protocol included unstable vital signs (heart rate >120 beats per minute, systolic blood pressure <90 mmHg [1 mmHg=0.133 kPa], respiratory rate >24 breaths per minute), ST-segment depressions or elevations, dynamic electrocardiogram (ECG) or new left bundle branch block (LBBB), troponin ≥ 0.06 ng/mL, persistent chest pain, pregnancy, acute altered mental status, or critical pertinent laboratory values.

CDU interventions

Upon admission to the CDU, patients received aspirin if not already given, received a chest X-ray if not done, telemetry monitoring, repeat troponin measurements and ECG every three hours (at least twice), and further testing, such as an exercise stress test or nuclear stress test. After monitoring and a minimum of two trended troponins three hours apart, a cardiologist evaluated the patient and then made the decision to perform provocative testing. Exercise stress testing was chosen for those under the age of 55 years, non-diabetics and capable of walking on a treadmill. Nuclear stress testing was chosen for all diabetic patients aged greater than 55 years and/or with two or more cardiac risk factors (current smoker, hypertension, hyperlipidemia, family history of cardiac disease). All patients were evaluated by a cardiologist prior to disposition. Patients were discharged home if there was improvement of clinical condition which included stable vital signs, no dynamic ECG changes or elevation in cardiac enzymes, normal stress test if obtained and follow-up in place. Patients were admitted if they had an abnormal stress test and were deemed to need cardiac catheterization, dynamic ECG changes, significant changes in cardiac enzymes, or unstable vital signs.

Statistical analysis

Statistical significance was identified by utilizing Chi-squared analysis, modified Fischer’s exact two-tailed t-test and multivariate logistic regression. Analyses were performed with SPSS version 26.0 statistical analysis software.

RESULTS

In all, 732 patients were admitted to the CDU, with 329 males (M) and 403 females (F). The age of the patients admitted to the CDU was 54.10±13.47 years for males (20–88 years, median 55 years) and 56.18±13.14 years for females (22–91 years, median 56 years). Males and females were found to have a median HEART score of 3 (M: 2.87±1.48; F: 2.85±1.28) (Table 1). The decision for further provocative testing was made in the direction of a cardiologist. There were no gender differences in the selection of patients for stress testing (M: 81.5% vs. F: 77.7%, χ²=1.5903, P=0.207) (Table 2). Gender differences were found in the rate of positive stress tests (M: OR=2.00, 95% CI: 1.39–2.86, P<0.001) (Table 3).

Table 1.

Demographics and descriptives

Table 2.

Gender comparison of admitted clinical decision unit patients undergoing stress testing

Table 3.

Logistical regression of biological sex with outcomes djusted by HEART score

Men were more likely to undergo catheterization than women, which correlates well with our findings that males were more likely to have positive provocative stress testing (M: OR=5.69, 95%CI: 2.89–11.19, P<0.001) (Table 3). Similarly, as provocative testing was more likely to be negative in females, so was the rate of disposition between genders. Males were more likely to have an admission to hospital from the CDU (M: OR=2.88 95%CI: 1.82–4.57, P<0.001) (Table 3). It is also worth noting that a majority of the patients admitted to the CDU were discharged (M: 79.6% and F: 92%; P<0.001), which correlates with the level of negative testing performed in both genders (Table 4). As expected, as HEART score increased, there was a stronger association with a positive stress test, cardiac catheterization requirement and the rate of admission (Table 3). Confidence intervals for the requirement of cardiac catheterization demonstrated wide ranges (Table 3); however, the trend of an association between elevated HEART score and the need for cardiac catheterization is observed. The wide range may be reflective of the low numbers of patients selected for cardiac catheterization.

Table 4.

Gender comparison of intervention and final disposition of clinical decision unit patients, n

DISCUSSION

In the case of low-risk chest pain and the absence of significant risk factors, the literature does not outline whether gender differences exist. In a Meta-analysis performed by Laureano-Phillips et al,[18] a low-risk HEART score had high sensitivity, negative predictive value, and negative likelihood ratio for predicting short-term major adverse cardiac events. This study demonstrates that there is a correlation between women with low-risk chest pain and a lower rate of positive provocative testing in comparison to men. Subsequently, women had a decreased need for further invasive testing and a higher rate of discharge home. Many studies have identified that men are more likely to have positive stress tests and significant coronary narrowing than women of the same age and risk.[19-21] In this study, men were more likely to have a positive stress test and undergo cardiac catheterization. The standard exercise stress test is recommended by the American Heart Association, the American College of Cardiologists, and the American Society of Nuclear Cardiology as the initial stress test of choice in most patients who can exercise, including women.[1,22,23] Exercise stress testing in women was thought to have decreased diagnostic accuracy because of a lower prevalence of CAD in women, but these early studies included almost exclusively male cohorts.[22, 24-26] At present, exercise stress testing with imaging or pharmacological testing is frequently the immediate choice of many physicians, without considering the important diagnostic and prognostic information that can be received from an exercise ECG.[1] Michaels et al[27] found that patients with a HEART score ≤3 being evaluated for chest pain in an observation unit are at extremely low risk for major adverse cardiac events and may be safely discharged without provocative testing. Positive cardiac testing in this population was more likely to represent a false-positive finding, resulting in unnecessary testing. Our findings are similar to those of Michaels et al.[27]

There are very few studies evaluating gender differences in patients with low-risk chest pain as stratified by a low HEART score. Our data indicate that in patients presenting with a low HEART score, gender may play a significant and independent role in the risk of ACS. This study shows that when adjusting for HEART score, there was a significant difference between gender and catheterization, implying that in the setting of similar cardiac risk factors at baseline, men are more likely to need cardiac catheterization and admission. Males were at least two times more likely than women to undergo catheterization in general and when adjusting for HEART score. After adjusting for HEART score, women were more likely to be discharged from the observation unit regardless of provocative testing findings. These findings suggest that HEART score alone may not be enough for risk stratification. This is also echoed by a recent methodological appraisal of the HEART score performed by Green et al.[28] They noted that the HEART score omits sex and other known predictors and has weak interrater reliability. They even go as far as to suggest that the widespread use of the HEART score and its variants should be reconsidered. [28]

Published studies have indicated that of patients admitted to the hospital for significant risk of ACS, women tend to have worse outcomes.[11, 14] This can only partly be explained by clinical differences where women admitted with ACS are usually older at presentation, have a higher burden of comorbidities, and tend to present later and with more atypical symptoms compared to men.[19, 20] Socioeconomic status, psychosocial burden and mental health have also been shown to be potential contributors to cardiovascular risk in women.[14] The HEART score was created as an objective means to assess cardiovascular risk in patients concerning ACS; however, gender is not utilized in the scoring rubric. Perhaps, the emphasis on atypical symptoms may explain the gender discrepancy that we are seeing in our low HEART score cohort. Although we did not evaluate this in our study, this hypothesis should be tested in a future prospective study.

This study also provides insight into the utilization of the CDU for low-risk chest pain, such that 86.3% of patients admitted to the CDU for ACS risk in 2017 were discharged after 24 h. The question arises whether these patients admitted to the CDU can be sent for outpatient follow-up as opposed to undergoing risk stratification testing in the ED. In hospitals that lack a CDU, a majority of these patients would be discharged home. The question now becomes whether the CDU should be utilized for higher-risk patients, such as those with higher HEART scores, to mitigate the burden of inpatient admission. However, further studies are required to stratify this patient population to ensure that those at higher risk for a cardiac event will be evaluated promptly while under observation.

Limitations

We recognize that this study was performed as a single-site retrospective qualitative chart review (RCR) of patients admitted to the observation unit. While the RCR is limited in its generalizability due to convenience sampling, it is still a popular methodology and can provide valuable information to direct subsequent prospective studies.[29] We also acknowledge that our study results are limited by the lack of post-discharge outcomes and that there is a high risk of bias and statistical heterogeneity in our study. However, our study does identify gender as a possible independent risk factor in the evaluation of patients admitted to the observation unit with concerns for ACS.

CONCLUSIONS

This study demonstrates that women with low-risk chest pain were less likely to have a positive stress test, to undergo cardiac catheterization, and to be admitted to the hospital from the CDU, than men of similar ages and HEART scores. While this raises the possibility of overemphasis of cardiac symptoms in women, it might also be reflective of the known bias of the atypical presentation of ACS in women.[9,12,14] Additionally, when we adjusted for HEART score, women were less likely to undergo cardiac catheterization and more likely to be discharged from the CDU, suggesting that gender could be an independent risk factor.