Outcomes of ST-Segment Elevation Myocardial Infarction in Patients With Adrenal Insufficiency

Abstract

Due to chronic treatment with exogenous glucocorticoids, patients with adrenal insufficiency (AI) have an increased risk of cardiovascular disease and cardiovascular mortality [1]. However, AI and chronic glucocorticoid treatment also leave patients vulnerable to adrenal steroid insufficiency during physiological stress [2]. Accordingly, when patients with AI present with acute cardiovascular events, they require supplementary steroid treatment.

Considering the associations between AI, cardiovascular disease, and inadequate stress response, we hypothesized that AI may predispose patients admitted with ST-segment elevation myocardial infarction (STEMI) to unfavorable outcomes during their hospitalizations. Limited data exist on STEMI outcomes when AI is present. Therefore, we queried a large national administrative claims database to characterize the outcomes of this patient cohort.

Materials and Methods

Data Source and Ethics Statement

Created by the Healthcare Cost and Utilization Project (HCUP) of the Agency for Healthcare Quality and Reporting (AHRQ), the National Inpatient Sample (NIS) database is the largest United States database for inpatient data. NIS is derived from billing data submitted by hospitals to HCUP, covering an estimated 97% of the US population [3, 4]. HCUP provides appropriate weighting variables, which extrapolate NIS data to provide national estimates. Because NIS data are deidentified and publicly available, this study was exempt from the requirements of the Hennepin Healthcare Institutional Review Board.

Study Population

The NIS database was queried from January 1, 2016, through December 31, 2019, to identify adult patients (≥18 years old) hospitalized with a primary diagnosis of STEMI (International Classification of Diseases, Tenth Revision, Clinical Modification [ICD-10-CM] codes I2101, I2102, I2109, I2111, I2119, I2121, I2129, and I213). Patients with a primary diagnosis of STEMI were divided into 2 groups based upon the presence or absence of an AI secondary diagnosis (ICD-10-CM E271, E273, E2740, E2749).

Baseline population demographic and comorbid characteristics were distinguished using appropriate ICD-10-CM codes which involved a family history of ischemic heart disease and personal history of hypertension, dyslipidemia, hypothyroidism, atrial fibrillation, chronic obstructive pulmonary disease (COPD), prior acute myocardial infarction, prior stroke, and diabetes mellitus. An entire list of the appropriate ICD-10 diagnosis and procedure codes used in this analysis is presented elsewhere (Table S1 [5]).

Study Outcomes

The primary outcome of interest was in-hospital mortality. Secondary outcomes included percutaneous coronary intervention (PCI), coronary artery bypass graft (CABG) intervention, acute kidney injury (AKI), ventricular tachycardia, need for vasopressors, need for mechanical ventilation, mechanical circulatory support (MCS) use, hospital length of stay (LOS), and total charges. Within the constraints of the NIS, total charges refer to the amount that hospitals billed payers for hospital services and exclude physician fees. No cost to charge conversion ratios were used in this study.

Statistical Analysis

Univariate logistic and linear regression analyses were used to calculate unadjusted odds ratios (uORs) for the primary and secondary outcomes. Multivariable logistic and linear regression analyses were subsequently used to adjust for the following confounders: sex, age, race, median household income by ZIP code, hospital characteristics (region, bed size, and teaching status), history of COPD, diabetes mellitus, and comorbidities for each outcome. Comorbidities adjustment used the 1993 Deyo adaptation of the Charlson Comorbidity Index, an adaptation of the original Charlson index, which can be calculated using NIS data. The latest Charlson Comorbidity Index includes the following 19 comorbid conditions: myocardial infarct, congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, connective tissue disease, ulcer disease, mild liver disease, diabetes, hemiplegia, moderate or several renal disease, diabetes with end organ damage, any tumor, leukemia, lymphoma, moderate or severe liver disease, metastatic solid tumor, and AIDS, each weighted based on their association with 1-year mortality [6]. Proportions were compared with chi-square tests, and continuous variables were compared with Student's t test. P < .05 was used as the threshold for statistical significance. All analyses in this study were performed using version 18.0 Stata (College Station, TX).

Results

Patient Characteristics

Of 691 430 patients admitted with a primary diagnosis of STEMI, 1382 (0.2%) had a secondary diagnosis of AI. Among patients with AI-STEMI, 87% had a diagnosis of secondary AI and 13% had a diagnosis of primary AI. Compared with patients without AI, patients with AI were older (mean age 66 years vs 63 years), were more likely to be female (45% vs 30%, P = .02), and more likely to have Medicare as a payer (65% vs 47%, P < .001). Patients with AI were more likely to have atrial fibrillation (20% vs 14% P < .001) and COPD (23% vs 11%, P < .001) but less likely to have hypertension (28% vs 48%, P < .001), dyslipidemia (51% vs 59%, P = .011), tobacco use (37% vs 46%, P = .007), and family history of ischemic heart disease (8% vs 15%, P = .01) compared with those without AI. No difference was observed in the prevalence of concomitant diabetes mellitus (33% vs 32%, P = .558). Baseline characteristics of patients with STEMI stratified by the presence or absence of AI are shown in Table 1 and Fig. 1.

Figure 1.

Bar graph illustration of the differences of baseline comorbidities between patients with vs without AI. HTN, hypertension; COPD, chronic obstructive pulmonary disease.

Open in new tabDownload slide

Primary and Secondary Outcomes

In the unadjusted analysis, AI was associated with higher odds of in-hospital mortality (uOR 2.3, 95% CI 1.62-3.26, P < .001). Additionally, patients with AI were less likely to receive PCI (66% vs 80%; uOR 0.50, 95% CI 0.38-0.66; P < .001) and more likely to received CABG (14% vs 5%; uOR 2.9, 95% CI 2.05-4.3; P < .001) during index hospitalization than patients without AI. Patients with AI also had higher odds of AKI (uOR 3.49, 95% CI 2.7-4.51, P < .001), ventricular tachycardia (uOR 1.71, CI 1.22-2.4, P = .002), need for vasopressors (5.4% vs 1.9% uOR 2.9, 95% 1.6-5.09 P < .001), mechanical ventilation use (28% vs 11% uOR 2.9, 95% 2.2-3.9 P < .001), and need for MCS (uOR 2.57, 95% CI 1.89-3.49, P < .001) compared with no AI. Unadjusted in-hospital outcomes are presented in Fig. 2.

Figure 2.

Bar graph illustration of unadjusted percentage outcomes of primary and secondary outcomes in patients with STEMI with vs without AI. CABG, coronary artery bypass graft; AKI, acute kidney injury; MCS, mechanical circulatory support; PCI, percutaneous coronary intervention.

Open in new tabDownload slide

After adjustment for baseline characteristics, AI remained associated with higher odds of in-hospital mortality (adjusted OR [aOR] 1.51, 95% CI 1.03-2.2, P = .033). AI also remained associated with lower odds of PCI (aOR 0.73, 95% CI 0.55-0.98, P = .040), higher odds of CABG intervention (aOR 2.8, 95% CI 1.89-4.2, P < .001), AKI (aOR 2.38, 95% CI 1.72-3.3, P < .001), ventricular tachycardia (aOR 1.55, 95% CI 1.1-2.2, P = .013), need for vasopressors (aOR 2.34, 95% CI 1.33-4.1 P = .003), need for mechanical ventilation (aOR 2.11, 95% CI 1.54-2.89 P < .001), and need for MCS (aOR 2.18, 95% CI 1.57-3.03, P < .001) compared with those without AI. Adjusted in-hospital outcomes are presented in Fig. 3.

Figure 3.

Forest plot demonstrating adjusted odds ratios (aORs) with 95% CIs of primary and secondary outcomes in patients with STEMI with vs without AI. PCI, percutaneous coronary intervention; AKI, acute kidney injury; MCS, mechanical circulatory support; CABG, coronary artery bypass graft.

Open in new tabDownload slide

LOS and Hospital Charges

Among STEMI hospitalizations, the average hospital LOS was 10 days for patients with AI and 4.02 days for patients without AI (P < .001). The adjusted hospital LOS was significantly longer in patients with AI than in those without AI (adjusted LOS increase of 5.78 days, 95% CI 4.23-7.5 days, P < .001). The average total charges were $258 475 for patients with AI and $115 505 for patients without AI. The adjusted total charges were significantly higher in patients with AI than in those without AI (adjusted charge increase $141,798, 95% CI $97,386-$198,554, P < .001).

Discussion

This study represents the first population-wide study relating prevalent AI to in-hospital outcomes in patients admitted with STEMI. The study demonstrates several novel findings: (1) patients with AI admitted for STEMI had distinctive demographic and clinical risk profiles compared to those without AI; (2) among patients admitted with STEMI, AI is associated with higher odds of in-hospital mortality, lower odds of undergoing PCI, higher odds of undergoing CABG, higher odds of AKI, need for vasopressors, mechanical ventilation, and MCS compared to those without AI; and (3) AI was associated with a longer mean LOS and higher total charges than patients without AI.

Demographic and Clinical Risk

Although the prevalence of AI in the general population is 10 to 15 per 100 000 [6], our study illustrated that 0.2% of patients (equivalent to 200 per 100 000) admitted with STEMI had AI, raising the possibility of an association between AI and ischemic heart disease. This finding is consistent with prior pathophysiologic and clinical observations. As patients with AI receive prolonged glucocorticoid therapy, occasionally with mineralocorticoids, they become vulnerable to both steroid insufficiency and excess [1, 7]. Glucocorticoid excess has been associated with an increased likelihood of developing obesity, hypertension, metabolic syndrome, and cardiomyopathies [8-10]. Conversely, acute adrenal crisis may result in acute cardiovascular complications [11, 12]. Esposito et al theorized that AI predisposes patients to lower blood pressure, which, in the setting of coronary artery disease, accentuates myocardial hypoperfusion [2].

Patients with AI in our study were older and more likely to be female. Additionally, patients with AI were less likely to have other traditional ischemic heart disease risk factors including hypertension, dyslipidemia, tobacco use, and family history of ischemic heart disease. This low prevalence of traditional cardiovascular risk factors suggests that their predisposition to worse in-hospital outcomes may result from noncardiac factors such as AI itself. Also, the AI population was predominantly female, and significant sex differences exist in ischemic heart disease risk profiles given higher rates of smoking, hypertension, and dyslipidemia in men than in women [13]. Uniquely, although diabetes is a major risk factor for heart disease and a potential side effect of steroid overreplacement, there was no difference noted between the 2 study groups in the prevalence of diabetes. Further studies are necessary to explore the demographic and clinical risk profile differences between patients with and without AI who present with STEMI.

In-Hospital Outcomes

Among patients hospitalized for STEMI, AI was associated with lower use of PCI than in those without AI. Patients with AI are known to have higher risks of procedural complications as they are prone to perioperative hypotension, respiratory depression, and prolonged responses to anesthesia [14]. Operators may shy away from PCI in these higher-risk patients. Furthermore, due to differences in pathophysiology, patients with AI-STEMI may not exhibit typical STEMI symptomatology, which could predispose them to a delayed presentation of their illness and consequently could delay their access to appropriate coronary intervention.

AI was associated with higher in-hospital STEMI mortality, AKI, and need for vasopressors, MCS, and mechanical ventilation interventions compared with those without AI. Patients with AI who suffer an acute event such as STEMI have an increased hormone demand and often require higher doses of steroid supplementation. This supply–demand mismatch can lead to acute relative steroid deficiency in patients with AI-STEMI, resulting in worsening hypotension and the need for vasopressors and MCS support. On the other hand, excessive steroid supplementation while inpatient may also result in worse in-hospital outcomes in patients with AI, with excess glucocorticoid supplementation associated with an increased risk of adverse cardiovascular outcomes [15]. Therefore, careful, and precise glucocorticoid dosing is necessary during STEMI hospitalizations to mitigate the risks associated with both low and high levels. Multidisciplinary care involving both cardiovascular medicine and endocrinology providers is crucial to optimize outcomes.

Patients with AI underwent higher rates of CABG interventions than those without AI. This could be attributable to a greater burden of multivessel coronary disease in patients with AI, given their increased propensity for accelerated atherosclerosis physiology [16]. AI was additionally associated with a higher incidence of peri-STEMI ventricular tachycardia than in those without AI. Serum electrolyte imbalances and acidosis in the setting of adrenal hormone deficiency could potentiate myocardial reperfusion injuries and impact the myocardial electrical environment, predisposing to rhythm disturbances. Iga et al suggest that catecholamine release induced by hypoglycemia might lead to arrhythmias and/or abnormal wall motion of the left ventricle in patients with AI [17]. Similarly, Nunoba et al identify hypomagnesemia in AI as shortening the effective refractory period and prolonging the relative refractory period of cardiac conductive tissue [18]. Several case reports have also previously described an increased propensity for ventricular arrhythmias in patients with AI [19, 20]. These complex physiologic and anatomic risks for patients with AI help explain the substantive increase in adverse STEMI outcomes including death.

LOS and Total Charges

Patients with AI hospitalized for STEMI had a longer hospital LOS and higher total charges than patients without AI. Although adverse outcomes such as the increased likelihood for vasopressors and mechanical ventilation could explain longer LOS and higher total charges in the AI group, the need for stress steroid management in patients with AI could further extend hospitalizations, as patients require additional care for AI in addition to STEMI. Steroid dose adjustments have been shown to prolong hospital LOS and increase demand for healthcare resources among hospitalized patients with AI [12].

Limitations and Strengths

The present findings should be interpreted in the context of limitations inherent to the NIS database. Details regarding mineralocorticoid and glucocorticoid treatments such as dosages were not available. Moreover, the NIS only captures inpatient data, and does not enable long-term posthospitalization follow-up and does not clarify if secondary diagnoses were present on admission or were diagnosed during the admission, and, hence, conclusions about duration of secondary diagnoses cannot be made [4]. Finally, ICD-10 codes are subject to error or incompleteness.

However, the AHRQ carries routine data quality assessments of the NIS to ensure data integrity which minimizes the likelihood of errors. Additionally when weighted, the NIS reflects the spectrum of the United States population. Furthermore, this is the first nationwide study to describe the association between AI and in-hospital outcomes among patients hospitalized for STEMI in the United States.

Conclusions

Patients with AI are admitted for STEMI at a rate disproportionate to their presence in the population, were less likely to receive PCI, more likely to receive CABG interventions, had worse in-hospital outcomes, and more frequent need for advanced support compared with patients without AI. Furthermore, AI was associated with longer LOS and higher total charges among patients with STEMI. While our findings call for heightened awareness of this high-risk population, the presence of limiting factors such as the inability to adjust for the etiology of AI and the inability to extrapolate and adjust for in-hospital steroid dosing regimens limits the formation of any clear causations attributable to such an association. It is therefore paramount that further studies be conducted to confirm these findings and to better clarify the reasons for such adverse hospital outcomes in patients with AI-STEMI.

Acknowledgments

We extend our sincere gratitude to Dr. Rosemary Quirk for her unwavering support of her residents, Dr. Pack Quinn for his generous time in reviewing this manuscript and offering insightful feedback and finally, to Dr. Mark Linzer—our mentor and enduring source of inspiration—none of this would have been possible without his unyielding guidance.

Funding

There was no funding allocated for this research.

Disclosures

A.M.G. reports speaking for Philips and Edwards Lifesciences and consulting for Philips and Inari Medical. H.D.A. is a Consultant for ReCor Medical, Medtronic, Philips, Silk Road Medical, BD. No other authors report relationships with industry to disclose.

Data Availability

The data underlying this manuscript are available in the Healthcare Cost and Utilization Project, specifically the Nationwide Inpatient Sample, accessible through the HCUP Central Distributor at: https://www.hcup-us.ahrq.gov/tech_assist/centdist.jsp. The datasets were derived from sources in the public domain: the NIS dataset is maintained by the Agency for Healthcare Research and Quality and is available to researchers upon request through the HCUP Central Distributor

References

Abbreviations

 
• AI

  adrenal insufficiency

 
• AKI

  acute kidney injury

 
• aOR

  adjusted odds ratio

 
• CABG

  coronary artery bypass graft

 
• COPD

  chronic obstructive pulmonary disease

 
• HCUP

  Healthcare Cost and Utilization Project

 
• ICD-10-CM

  The International Classification of Diseases 10th Revision Clinical Modification

 
• LOS

  length of stay

 
• MCS

  mechanical circulatory support

 
• PCI

  percutaneous coronary intervention

 
• NIS

  Nationwide Inpatient Sample

 
• STEMI

  ST-segment elevation myocardial infarction

 
• uOR

  unadjusted odds ratio

 
• VT

  ventricular tachycardia