Disparities in Cardiovascular Outcomes Among Emergency Department Patients with Mental Illness

Abstract

BACKGROUND:

Patients with mental illness have been shown to receive lower quality of care and experience worse cardiovascular (CV) outcomes compared to those without mental illness. This present study examined mental health-related disparities in CV outcomes after an Emergency Department (ED) visit for chest pain.

METHODS:

This retrospective cohort included adult Medicaid beneficiaries in Washington state discharged from the ED with a primary diagnosis of unspecified chest pain in 2010-2017. Outcomes for patients with any mental illness (any mental health diagnosis or mental-health specific service use within 1 year of an index ED visit) and serious mental illness (at least two claims (on different dates of service) within 1 year of an index ED visit with a diagnosis of schizophrenia, other psychotic disorder, or major mood disorder) were compared to those of patients without mental illness. Our outcomes of interest were the incidence of major adverse cardiac events (MACE) within 30 days and 6 months of discharge of their ED visit, defined as a composite of death, acute myocardial infarction (AMI), CV rehospitalization, or revascularization. Secondary outcomes included cardiovascular diagnostic testing (diagnostic angiography, stress testing, echocardiography, and coronary computed tomography (CT) angiography) rates within 30 days of ED discharge. Only treat-and-release visits were included for outcomes assessment. Hierarchical logistic random effects regression models assessed the association between mental illness and the outcomes of interest, controlling for age, gender, race, ethnicity, Elixhauser comorbidities, and health care use in the past year, as well as fixed year effects.

RESULTS:

There were 98,812 treat-and-release ED visits in our dataset. At 30 days, enrollees with any mental illness had no differences in rates of MACE (AOR 0.96; 95% CI, 0.72-1.27) or any of the individual components. At 6 months, enrollees with any mental illness (AOR 1.86; 95% CI, 1.11-3.09) and serious mental illness (AOR 2.60; 95% CI 1.33-5.13) were significantly more likely to be hospitalized for a CV condition compared to those without mental illness. Individuals with any mental illness had higher rates of testing at 30 days (AOR 1.16; 95% CI 1.07-1.27).

CONCLUSION:

Patients with mental illness have similar rates of MACE, but higher rates of certain CV outcomes, such as CV hospitalization and diagnostic testing, after an ED visit for chest pain.

1.1. INTRODUCTION

Patients with mental illness experience poorer health outcomes including higher rates of cardiovascular morbidity and mortality compared to those without mental illness.[1–6] Having a serious mental illness has been associated with a 2-fold higher risk of death and nearly 25 years of life lost, with the majority of these premature deaths arising secondary to cardiovascular disease.[3, 6–7] While much of this disparity is related to concomitant socioeconomic factors and higher rates of risk factors like smoking and obesity,[8] evidence suggests that patients with mental illness may sometimes receive lower quality of care from health care professionals.[9] For example, rates of cardiac catheterization, revascularization, and reperfusion therapies following an acute myocardial infarction (AMI) have been shown to be lower for patients with comorbid mental disorders.[10–11] Patients with mental illness have a higher mortality post-AMI that may be explained by differences in the provision of evidence-based care, though the literature is mixed.[12–13] In the emergency department (ED), a charted history of depression is independently associated with receiving lower-priority triage scores and experiencing more delays in diagnostic testing and reperfusion when having an AMI.[14] Finally, some studies have suggested that patients with mental illness receive preventative services for cardiovascular disease, such as cholesterol monitoring, at a lower rate when compared to those without mental illness.[15–16]

In the past decade, ED visits among patients with mental illness have increased substantially, especially in the Medicaid population.[17] State Medicaid programs are the largest payers of mental health services in the United States, providing coverage for nearly a third of US adults with serious mental illness.[18] Yet, to date, few studies have examined the relationship between co-existing mental illness and healthcare outcomes in the ED in the Medicaid population. The goal of this project was to examine disparities in cardiovascular outcomes among Medicaid beneficiaries with mental illness who present to the ED for chest pain. Using Washington state Medicaid claims, we compared cardiovascular outcomes and diagnostic testing rates for beneficiaries with and without mental illness.

1.2. METHODS

1.2.1. Data Source and Study Population

We conducted a retrospective study of adult Medicaid beneficiaries in Washington state who had a treat-and-release ED visit for chest pain between 2010-2017. We excluded children <18 years of age, beneficiaries who were dually enrolled in Medicaid and Medicare, beneficiaries who were not enrolled for at least 7 of 12 months preceding their index visit (to allow for churn in enrollment but adequate capture of comorbidities for risk adjustment), and beneficiaries who were not continuously enrolled for at least 6 months after their index visit (to allow for outcomes assessment).

ED visits were identified from outpatient facility and professional claims by a CPT code of 99281-99825, a revenue code of 0450-0459, or a place of service listed as the ED. We included visits with either an International Classification of Diseases (ICD) - 9th revision diagnosis of 786.5x for claims before October 2015, or ICD-10th revision diagnosis of R07.89-9 for claims October 2015 and later, listed as the primary diagnosis. To ensure capture of unique episodes of care, we further excluded index ED visits in which the patient had an ED visit within the past 30 days or 6 months (for 30-day or 6-month outcomes, respectively).

1.2.2. Identification of mental illness

We adapted the Washington state Department of Social and Health Services definition of mental health service need,[19] defining mental illness as any mental health diagnosis or mental-health specific service use within 1 year of an index ED visit (a 24-month identification window, which for this study is anchored on the index visit). As this definition is designed to be broad and includes those with those with both mild-to-moderate as well as serious mental health conditions, we also examined outcomes for a subset of patients with serious mental illness. To identify this subgroup, we used existing definitions for severe and persistent mental illness, which requires at least two claims (on different dates of service) within 1 year of an index ED visit with a diagnosis of schizophrenia, other psychotic disorder, or a major mood disorder.[20]

1.2.3. Outcomes

Our primary outcome was the incidence of major cardiac adverse event (MACE) within 30 days of ED discharge. Only treat-and-release ED visits were considered for outcomes analysis, though we looked at disposition from the ED for the entire sample. We defined MACE as a composite of death, revascularization, acute myocardial infarction, or any cardiovascular-related readmission. Any inpatient admission with a relevant cardiovascular primary diagnosis within the 30-day window, regardless of the source of admission (ED or direct admission), was counted as a cardiovascular hospitalization. We used a combination of ICD-9 and ICD-10 diagnosis and procedure codes, as well as CPT codes, to identify relevant claims (Appendix Table A1). Secondary outcomes included cardiovascular diagnostic testing rates within 30 days of ED discharge (including diagnostic angiography, stress testing, echocardiography, and coronary computed tomography (CT) angiography). We also examined MACE within 6 months of ED discharge.

1.2.4. Statistical Analysis

Demographic and clinical characteristics of chest pain ED visits for Medicaid enrollees with any mental illness and serious mental illness were compared to those without mental illness using standard bivariate analyses. We employed hierarchical logistic random effects regression models to assess the association between mental illness and the outcomes of interest following discharge from the ED for chest pain, controlling for clustering within hospitals. Models controlled for age, gender, race, Elixhauser comorbidities,[21] and health care use in the past year (including ED visits and inpatient hospitalizations), as well as indicators for year. All analyses were completed using Stata version 16 (StataCorp).

1.3. RESULTS

1.3.1. Characteristics of Visits

We identified 116,814 ED visits for nonspecific chest pain, of which 98,812 visits (84.6%) were treat-and-release visits. Mental health conditions were common in this sample of Medicaid ED users. Enrollees with any mental illness comprised 69.2% of all and 74.7% of the treat-and-release visits for chest pain. A quarter of all chest pain visits (and 30.0% of treat-and-release visits) were among enrollees meeting criteria for having a serious mental illness. The proportion of ED visits among enrollees with any mental illness or serious mental illness declined over time (Figure 1).

Figure 1.

Proportion of chest pain ED visits that were made by individuals with a mental illness.

Characteristics of the subset of treat-and-release visits, which were used for outcomes analysis, are described in Table 1. Most (88%) of the visits occurred in urban hospitals. On average, ED visits among enrollees with mental illness were more likely to be female and white, with a higher burden of comorbid conditions (including substance use) and higher rates of ED use in the past year compared to those without mental illness. These differences were most pronounced among those with serious mental illness.

Table 1.

Characteristics of Medicaid treat-and-release visits for chest pain, stratified by presence of mental illness

	No Mental Illness	Any Mental Illness	Serious Mental Illness
	N=25,025	N=73,787	N=29,618
Age, years, mean (SD)	39.0 (13.6)	40.9 (12.3)	41.1 (0.06)
Female, %	56.9	63.6	63.7
Race, %
White	51.7	64.8	67.0
Black/African American	13.9	11.8	13.4
Latinx	15.2	10.0	6.8
Asian/Pacific Islander	6.7	2.3	1.7
Native American/Hawaiian	5.3	5.2	5.0
Other	3.6	2.8	2.7
Unknown	3.6	3.1	3.5
Comorbidities, mean (SD)	1.7 (2.0)	4.3 (3.1)	5.3 (3.2)
Substance use, %	20.1	51.5	64.7
Healthcare Use, past 1y, mean (SD)
ED visits	3.1 (4.2)	8.5 (11.7)	10.9 (13.8)
Inpatient Admissions	1.0 (0.5)	1.2 (1.2)	1.2 (1.2)
Hospital Setting, %
Small Town/Rural	3.5	3.4	2.9
Large Town (10-50K)	9.3	8.5	6.7
Urban/Suburban	87.2	88.2	90.4

1.3.2. Outcomes Analysis

Enrollees with any mental illness had significantly higher rates of hospitalization during their ED visit than those without mental illness (16.0% vs 13.7%; AOR 1.15; 95% CI 1.10-1.22)). Overall hospitalization rates for those with serious mental illness were not statistically significantly different than those without mental illness (14.2% vs. 13.7%; AOR 1.07; 95% CI 0.98-1.18). Enrollees with mental illness who were discharged from the ED after an evaluation for chest pain did not experience significantly worse cardiovascular outcomes at 30 days compared to those without mental illness. Overall, MACE were rare in our study population. Only 338 (0.46%) ED treat-and-release visits among patients with any mental illness and 77 (0.31%) visits among those without mental illness experienced a MACE at 30-days. After adjusting for case-mix, these differences were not statistically significant for the composite outcome (AOR 0.96; 95% CI, 0.72-1.27) nor for any of the individual components (revascularization, cardiovascular hospitalization, AMI) in the 30-day MACE measure (Table 2). At 6 months, enrollees with any mental illness (0.67% vs. 0.19%; AOR 1.86; 95% CI, 1.11-3.09) and the subgroup with serious mental illness (AOR 2.60; 95% CI 1.33-5.13) were significantly more likely to be hospitalized for a cardiovascular condition compared to those without mental illness; however, overall composite MACE at 6 months was similar to those without mental illness.

Table 2.

Cardiovascular (CV) outcomes among patients with and without mental illness.

	No Mental Illness	Any Mental Illness	AOR (95% CI)	Serious Mental Illness	AOR (95% CI)
Any Hospitalization	3,969 (13.7%)	14,033 (16.0%)	1.15 (1.10, 1.22)	4,906 (14.2%)	1.07 (0.98, 1.18)
Observation Stay	693 (2.4%)	2535 (2.90%)	1.13 (1.04, 1.24)	903 (2.6%)	1.31 (1.15, 1.50)
Inpatient Admission	3,276 (11.3%)	11,498 (13.1%)	1.14 (1.08, 1.21)	4,003 (11.6%)	1.02 (0.93, 1.12)
MACE at 30 Days	77 (0.31%)	338 (0.46%)	0.96 (0.72, 1.27)	130 (0.44%)	0.79 (0.45, 1.39)
Revascularization	141 (0.49%)	369 (0.42%)	0.84 (0.59, 1.19)	92 (0.27%)	0.49 (0.23, 1.07) 0.51 (0.23, 1.10)
CV Hospitalization	12 (0.05%)	143 (0.19%)	1.68 (0.72, 3.90)	55 (0.19%)	1.74 (0.57, 5.31)
AMI	20 (0.08%)	101 (0.14%)	1.28 (0.67, 2.41)	47 (0.16%)	0.91 (0.26, 3.18)
MACE at 6 Months	189 (0.76%)	1077 (1.46%)	1.18 (0.94, 1.48)	409 (1.38%)	1.29 (0.88, 1.90)
Revascularization	143 (0.72%)	1062 (1.21%)	1.00 (0.77, 1.30)	322 (0.93%)	0.63 (0.37, 1.10)
CV Hospitalization	48 (0.19%)	498 (0.67%)	1.86 (1.11, 3.09)	190 (0.64%)	2.60 (1.33, 5.13)
AMI	48 (0.19%)	329 (0.45%)	1.15 (0.81, 2.97)	157 (0.53%)	1.41 (0.67, 2.97)
Diagnostic Testing at 30 Days	1,835 (6.33%)	6,975 (7.94%)	1.16 (1.07, 1.27)	2,512 (7.28%)	1.08 (0.91, 1.27)
Diagnostic Angiogram	177 (0.20%)	924 (1.05%)	1.19 (0.91, 1.55)	295 (0.85%)	1.08 (0.67,1.75)
Stress Test	996 (3.44%)	3,346 (3.81%)	1.19 (1.05, 1.36)	1185 (3.43%)	1.09 (0.87, 1.37)
Echocardiogram	1014 (3.50%)	4,061 (4.62%)	1.13 (1.00, 1.28)	1,507 (4.37%)	1.08 (0.88, 1.33)
CCTA	42 (0.14%)	177 (0.20%)	0.82 (0.53, 1.25)	83 (0.24%)	1.24 (0.59, 2.62)

Notes: The assessment of 30-day and 6-month outcomes was conducted on the subset of treat-and-release ED visits.

While only a small number of patients went on to complete any diagnostic testing within 30-days of discharge, individuals with any mental illness had higher rates of testing at 30 days (7.94% vs. 6.33%, AOR 1.16; 95% CI 1.07-1.27), including completing significantly more stress tests (AOR 1.19; 95% CI 1.05-1.36) and echocardiograms (AOR 1.13; 95% CI 1.00-1.28) compared to those without mental illness. Among the serious mental illness subgroup, diagnostic testing rates were not statistically significantly different from those without mental illness (AOR 1.08; 95% CI 0.91-1.27). There was no difference in the proportion of individuals with and without mental illness who underwent diagnostic angiogram or coronary CT.

1.4. DISCUSSION

In this retrospective study of Medicaid beneficiaries, we examined whether patients with mental illness experienced disparities in cardiovascular outcomes and diagnostic testing after an ED visit for chest pain. Our study finds no evidence that patients with mental illness undergoing evaluation for chest pain in the ED experience disparate emergency care when compared to those without mental illness. We found that patients with mental illness were more likely to be hospitalized during their index visit, and at 30 days had similar outcomes and completed more cardiac diagnostic testing compared to those without mental illness. Higher rates of cardiac testing, stress tests and echocardiograms in particular, were observed when examining all patients with mental health conditions, but not in the subgroup with serious mental illness. In this latter group, there were no differences in 30-day diagnostic testing or outcomes. Overall, only a very small number of ED visits in our population, amounting to ≤ 0.5% of total discharges of those with and without mental illness, had a MACE at 30 days. These results are consistent with event rates in other studies of patients discharged from the ED after an evaluation for chest pain,[22–24] and highlight the overall safety of ED dispositions for chest pain, even among higher-risk, vulnerable groups.

We did find that the presence of mental illness was associated with a higher rate of cardiovascular hospitalization at 6 months. While patients with mental illness also had higher rates of AMI at 6 months, the differences did not rise to statistical significance. These higher rates of hospitalization comprise of a variety of cardiovascular conditions and could be related to greater cardiac disease burden,[1-6] barriers to accessing healthcare services,[25-26] and other poor social determinants of health that impact the management of chronic disease.[27-28] In our sample, Medicaid enrollees with mental illness had a substantially greater burden of chronic comorbidities compared to those without mental illness, which likely contribute to the higher rates of cardiovascular hospitalization. Moreover, substance use was more than 2.5 times as prevalent in the population with any mental illness, and more than 3 times as prevalent in the population with severe mental illness compared to those without mental illness. While our models adjusted for co-existing substance use, this differential rate of substance abuse likely remains an important factor related to the increased cardiovascular disease burden and hospitalizations. Strategies like care coordination and integrated physical-behavioral health service delivery in the ambulatory setting have been shown to improve quality of life and outcomes, and could be more effective at preventing bounce-back ED visits than changes to ED management of these patients.[29-31]

While our claims-based analysis does not allow us to study all aspects of care delivery, our study suggests that patients with mental illness likely receive similar workups and are hospitalized at a slightly higher rate, consistent with their higher baseline risk for cardiovascular disease. This may be a function of an increasingly standardized approach to the ED evaluation of chest pain over time, which employs the use of clinical decision rules like the HEART score in conjunction with serial contemporary troponin testing, that has been shown to miss very few cases of acute coronary syndrome and accurately risk-stratifies patients for disposition.[32-34] It is possible that disparities in outcomes between those with and without mental illness would be more pronounced for other cardiovascular conditions which involve greater physician discretion and less formalized treatment algorithms (for example, heart failure), though that was not the purpose of the present study.

Overall, completion of stress and other cardiac testing was markedly lower in our Medicaid population compared to that reported in other studies of ED patients with chest pain.[35-36] Low rates of cardiac testing at 30 days could reflect challenges experienced by Medicaid patients accessing timely follow-up care or scheduling health care appointments.[37] Among the Medicaid enrollees in our sample, it appears that individuals with less serious mental illness are more likely to receive testing in the 30-days after ED discharge, whereas we observed no difference in diagnostic testing among the subgroup with serious mental illness.

Identifying the underlying etiology of differences in testing and treatment patterns is complicated and may reflect both patient and provider characteristics. There is a reasonable body of literature suggesting that many patients who present to the ED for acute chest pain and are ultimately ruled out for AMI have symptoms consistent with anxiety.[38-40] Thus, differential rates in 30-day testing may reflect a higher propensity to seek care among this population. However, differences could also be driven by provider bias around mental illness. For example, providers could attribute cardiac symptoms incorrectly to anxiety simply because a patient has a known history of mental illness. In one recent study, 1.3% of AMIs were missed by ED providers, with a higher odds of missed diagnosis among individuals with mental illness.[41] As a result, there may be cases in which providers may be somewhat less likely to refer certain patients for cardiac evaluations, though overall rates of MACE remain low.

Our study has several important limitations. First, this analysis of claims data relies on discharge diagnoses to identify our cohort and assess outcomes among those patients discharged after their ED evaluation. This means that patients have already been determined to be lower-risk at the time we follow them for outcomes. It could also mean that we miss potential adverse outcomes in individuals with mental illness who have chest pain misdiagnosed as something else, such as anxiety. Had we been able to follow patients based on chief complaint through their ED encounter, other disparities in triage or delays in evaluation may have emerged. Second, our population of Medicaid beneficiaries had especially high rates of mental illness, wherein nearly three-quarters of ED visits for chest pain in our sample were made by patients with a mental health condition, and thus might not be generalizable to all ED patients with chest pain. Third, we classify mental illness based on any mental health diagnosis or service use within +/−1 year of an index ED visit, which means that some individuals in our mental illness cohort may rarely be diagnosed with mental illness after their ED visit. While this criteria is meant to be broad and identify the maximal number of individuals with a mental health condition, we recognize that the choice of identification window may change the observed associations, especially given that many outcomes are rare. Finally, we only studied chest pain, but other diagnoses that do not have clear testing and treatment algorithms, like heart failure, may show greater disparities in care.

1.4.4. Conclusion

In this study, patients with mental illness have similar rates of MACE, including AMI after ED discharge, but higher rates of certain cardiovascular outcomes, such as cardiovascular hospitalization and diagnostic testing, after an ED visit for chest pain. Further research should be conducted to better elucidate potential disparities in the management and outcomes of patients with mental illness in the ED setting and investigate potential reasons, and solutions, for such findings.