INTRODUCTION
Acute decompensated heart failure (ADHF) is the leading cause for hospitalization in the United States among adults ≥80 years of age and the second most common cause among adults ≥65 years [1]. Although medications are effective for controlling ADHF symptoms, exacerbation of symptoms is common and frequently leads to emergency department (ED) encounters. Every year, over one million ED encounters in the United States are related to ADHF, the majority of which require admission (approximately 80%) [2–4]. Identification and treatment of ADHF in the ED is complex, partly due to the heterogeneous nature of ADHF and variability in presenting symptoms. This variability leads to a range of diagnostic options for ED physicians (e.g., echocardiogram, chest x-ray, B-type natriuretic peptide [BNP], ultrasound), leading to time wasted ordering and collecting data [5].
A better understanding of sex and age differences in ADHF presentation in the ED could increase the efficiency and accuracy of ADHF diagnosis, with the potential to reduce discrepancies in care and improve patient outcomes and quality of life [6]. ADHF may present with a plethora of complaints, some more typical than others. Shortness of breath, lower extremity edema, and chest pain are considered the most “typical” symptoms. ADHF may also present with “atypical” symptoms such as confusion, weakness, nausea, and vomiting [7]. Women hospitalized with ADHF are more likely to be older, have nonischemic causes of ADHF, present with new-onset ADHF, have higher blood pressure, and have orthopnea than men with ADHF [7]. Evidence also suggests that women are treated less intensively and hospitalized longer and more often than men [7,8]. Age differences have also been considered. Younger patients with ADHF are more likely to be obese, have more comorbidities, and have worse quality of life, whereas older patients are more likely to have atrial fibrillation and higher mortality rate [9]. No studies, however, have investigated sex and age differences in presenting chief complaints among ADHF patients in the ED.
Our aim in this study was to evaluate sex and age differences in ADHF symptom presentation in the ED. Characterizing symptom presentation of ADHF patients in the ED may help improve clinical care for ED patients with ADHF. We used statewide syndromic surveillance data to describe sex and age differences in chief complaints of ED patients with ADHF diagnoses [10].
METHODS
Ethics statement
The Institutional Review Board of The University of North Carolina at Chapel Hill determined that this study was exempt from approval (No. 17-1271). Informed consent was waived due to the retrospective nature of the study.
Study population
We conducted a retrospective analysis of ED encounters for patients ≥18 years old with ED diagnoses of ADHF using the North Carolina Disease Event Tracking and Epidemiologic Collection Tool (NC DETECT), a syndromic surveillance system containing state mandated information for all civilian ED encounters across the state [10]. For this study, we obtained patient age group, patient sex, ED characteristics, chief complaint and other intake details, and discharge diagnosis codes from NC DETECT. The study population included adult patient encounters between 2010 and 2016 with International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) and ICD-10-CM diagnosis codes for ADHF. ADHF visits were defined using ICD-9-CM discharge diagnosis codes 410–414, 427.5, and 428–428.9 in any discharge code position between January 1, 2010, and September 30, 2015. For data collected in 2016, ADHF encounters were defined by patients with ICD-10-CM discharge diagnosis codes I20–I25, I46, and I50–I50.9 in any position. Due to the transition from ICD-9-CM and ICD-10-CM in 2015, we excluded data from October 1, 2015, to December 31, 2015, from analyses.
Measures
Predefined age groups in NC DETECT included patients 18–24, 25–34, 35–44, 45–54, 55–64, 65–74, and ≥75 years of age. Biological sex was denoted as male, female, or unknown based on NC DETECT data. Race and ethnicity were not included in our analysis since they were not captured in NC DETECT before 2016. ED characteristics included hospital type (academic, nonacademic, and standalone ED) and location (coastal, piedmont, and mountain). Comorbid diagnoses were gathered using ICD-9-CM and ICD-10-CM codes and stratified into patient age groups of 18–44, 45–64, and ≥65 years.
Chief complaint categories
ED chief complaints provided by local EDs as free or descriptive text are standardized in NC DETECT using Emergency Medical Text Processor (EMT-P), a natural language processing software package [8]. Not all free-text phrases were successfully translated by the software. Project team members (JJB, MLM, MB, and TFPM) followed a three-level category system to classify EMT-P standardized chief complaints or free-text chief complaints consisting of 50 detailed symptom categories (level 1) related to ADHF, 29 rollup categories (level 2), and 12 body system categories (level 3) when standardization was not provided by NC DETECT (Supplementary Table 1). A chief complaint of hypoxia, for example, was categorized as a breathing issue at the most detailed level (level 1), breathing issues were rolled up into a respiratory category (level 2), and finally categorized as pulmonary complaints at the body system level (level 3). Categorizations were programmatically assigned by one investigator, reviewed for completeness and validity by a second investigator, and finally verified by a third, with adjudication. Chief complaints could map to more than one detailed level (level 1); however, rollup categories (level 2) and body system (level 3) categories are mutually exclusive. The level 1 category “other respiratory” includes symptoms of cough and wheezing.
Statistical analysis
Descriptive statistics were used to summarize chief complaints, comorbid conditions, and patient characteristics of ED encounters documented by NC DETECT. Stratified analyses of chief complaints for ED encounters were summarized according to sex and age (18–44, 45–64, ≥65 years). Comorbid conditions were stratified only by age (18–44, 45–64, ≥65 years). Standardized differences between groups were evaluated and considered to be meaningful at a comparative difference threshold of 10%, consistent with medical literature standards [11,12]. Since more than one condition could be reported as a chief complaint, patients could be summarized more than once per visit. However, encounters were only summarized once per category level. Data analyses were completed using SAS ver. 9.4 (SAS Institute Inc).
RESULTS
In total, we included 893,950 unique ED encounters with an ICD-9-CM or ICD-10-CM code for ADHF in the study. These diagnoses encompassed 422,720 unique patients, with a mean visit number of 1.6±1.3 and with 10,301 patients (1.8%) having more than five visits. At the time of their encounters, 578,313 patients (64.7%) were age 65 years or over, 260,397 (29.1%) were 45–64 years, and 55,216 (6.2%) were age 18–44 years; 491,868 ED encounters (55.0%) were female (Table 1). No differences in disposition were seen for chief complaints when stratified by between age and sex (data not shown). The top chief complaint categories were dyspnea (19.1%), chest pain (13.5%), and other respiratory complaints (13.4%) (Table 2).
Differences in chief complaint according to sex
There were no differences in ED chief complaints when stratified by sex alone (Table 2). Top presenting complaints were consistent for both sexes, the most common being dyspnea (20.1% in men, 18.2% in women), other respiratory (13.4% in men, 13.5% in women), and chest pain (14.7% in men, 12.5% in women).
When stratified by both age and sex, however, there was a meaningful standardized difference within the youngest age group (Fig. 1). Among patients 18–44 years old, women had more reports of nausea/vomiting compared with men (6.7% vs. 4.1%; standardized difference, 0.11) and headache compared with men (4.2% vs. 2.0%; standardized difference, 0.13). No difference was observed between women and men in the 45–64 years age group.
Differences in chief complaint according to age
When stratified by age, there were meaningful standardized differences in chief complaints between the 18–44 and 45–64 years age groups to the ≥65 years age group (Table 3). When compared with the ≥65 years group, both the 18–44 and 45–64 years age groups were more likely to present with chief complaints of chest pain and less likely to present with chief complaints of balance, weakness, and confusion. Additionally, when compared with the ≥65 years group, the 18–44 years group had more chief complaints of headache (standardized difference, 0.17), nausea/vomiting (standardized difference, 0.12), unclassified (standardized difference, 0.14), and fewer chief complaints of other respiratory problems (standardized difference, –0.12).
Comorbid conditions
In addition to chief complaints, we conducted descriptive analyses of the most prevalent comorbid diagnoses with heart failure according to age (Fig. 2 and Supplementary Table 2). The most common diagnoses occurring with ADHF diagnosis were essential hypertension (27.9%), diabetes mellitus (DM) (26.4%), and cardiac dysrhythmias (19.9%). Across all age groups, DM and hypertension were among the top 3 diagnoses associated with ADHF. However, cardiac dysrhythmias, new-onset or preexisting, only appeared in the top 5 associated diagnoses in patients ≥65 years and older, among whom it was the second most frequently associated diagnosis for individuals 65+ (15.6%), and the top associated diagnosis for individuals ≥75 years old (data not shown).
DISCUSSION
Within a statewide cohort of ADHF patients reporting to the ED, we found minimal differences in presenting chief complaints according to sex, meaningful standardized differences between the oldest (≥65 years old) patients and other age groups, and that ED co-diagnoses varied significantly with age. To our knowledge, this is the first study exclusively evaluating differences in chief complaints according to sex and age among ADHF patients presenting to the ED.
We observed a higher-than-expected percentage of atypical ADHF symptoms, with less than 20% of encounters listing shortness of breath (i.e., dyspnea) as their chief complaint, although this is considered the most conventional symptom for ADHF. Other symptoms classically associated with ADHF such as edema and respiratory complaints other than dyspnea were reported by 3.7% and 14.2% of our study population, respectively. Together, these three classic symptoms comprised less than 40% of all chief complaints in our cohort of ED patients presenting with ADHF. This indicates that ADHF patients commonly present to the ED with atypical symptoms.
Sex and age patterns in our study are similar to those of other studies, in which younger patients with ADHF were more likely to be male and older patients were more likely to be female [8,13,14]. However, we did not observe significant differences in ED chief complaints according to sex. We detected differences in chief complaints according to sex only in the youngest age through stratified analysis. Women who were 18–44 years of age presented more frequently with nausea/vomiting and headaches than 18–44 years old men, while no difference was observed between sexes in the 45–64 and ≥65 years age groups. This aligns with Sethares and Chin [15], who observed an increased prevalence of nausea in women. Our results differ, however, as that previous study demonstrated a higher incidence of fluid overload states (weight gain, edema) in male ADHF patients than in women. Meanwhile, Galvao et al. [16] demonstrated that women are less likely to receive aggressive treatment (vasoactive medication, procedure-oriented therapy) but have similar outcomes as men (length of stay, risk-adjusted in-hospital mortality), a finding that aligns with those of the present study. A significantly greater number of women are hospitalized due to heart failure with persevered ejection fraction, and significantly elevated BNP level has been linked to worse outcomes [17]. However, no significant differences between sexes were seen in long term outcomes in previous studies [16,17]. In the present study, we did not distinguish between ADHF with preserved verses reduced ejection fraction. Our findings reinforce previous observations of minimal differences in presentation, disposition, and outcomes by sex in ADHF patients.
Effects of age on chief presenting complaints were more common in younger than older age groups. Older patients with heart failure tend to show greater adherence to treatment regimens than younger patients [18], which could lead to better management of typical symptoms in ADHF, resulting in initial presentation with atypical symptoms. While poor medication adherence is correlated with increased risk of adverse cardiac events, mortality increases with age [18]. Despite having fewer classic comorbid conditions with ADHF, older patients were fourfold more likely to experience in-hospital cardiac death than younger patients [19]. Like the present study, Sethares and Chin [15] also stratified symptoms by age but observed a higher prevalence of chest pain in older individuals, countering our findings. The mean age of patients in that study, however, was 76.9 years, a much older population than the present study. Differences in chief complaint among ED patients according to sex and age have not been well characterized previously.
Our analysis of comorbidities with ADHF exacerbations aligns with current knowledge by demonstrating differences according to age and chief complaints [5]. Younger patients with ADHF tend to have greater incidence of obesity and DM, while older patients have greater incidence of coronary artery disease and hypertension [14,20]. Previous research demonstrated declines in the prevalence of ischemic heart disease, previous myocardial infarction, hypertension, DM, and chronic obstructive pulmonary disease among those presenting to the ED with ADHF after 80 years of age [14,21]. Additionally, Tisminetzky et al. [9] showed that patients presenting with a first acute myocardial infarction have differences in chief complaints according to age, similar to ED patients with ADHF. Meanwhile, atypical presentations of acute myocardial infarction and ADHF increase with age. Previous research and our findings agree that younger patients who experience ADHF are more likely to suffer from comorbidities that can be treated through lifestyle modifications and medication compliance than older patients. Obesity in younger adults is more likely to precipitate comorbid conditions earlier in life, including hypertension and DM, leading to a higher incidence of ADHF [22]. Older patients presenting with ADHF suffer more frequently from age-related degenerative heart disease symptoms, such as arrhythmias, which are not common contributors to ADHF. Likewise, presentations of ADHF follow a similar pattern, with varying comorbidities for different age groups [22,23]. Younger patients with ADHF exacerbations present with more typical symptoms, most likely due to poor compliance with treatment for other chronic conditions [20]. Older patients with ADHF tend to present with atypical symptoms consistent with natural age-related changes or infections [2].
There are several limitations to our work. All data used in this study were collected from 2010–2016 and analyzed retrospectively. As with any retrospective study requiring data categorization, our data are therefore subject to misclassification and inherent bias. However, in the future we intend to extend this project to include data collected after the acute phase of the COVID-19 pandemic, as new-onset ADHF has been associated with COVID-19 diagnosis in ED visits [19]. Because this was a large study with nearly 900,000 data points, limited information was collected for each patient. No data were available regarding socioeconomic status or ED setting (rural or urban), and race/ethnicity was not available for 79% of the encounters, so this information was not reported. This is a limitation that will be addressed in future studies. The study population represents only ED visits in North Carolina and may not be generalizable to populations in other states or countries. Data were collected using both ICD-9-CM and ICD-10-CM codes, as a transition in diagnostic codes occurred in 2015. However, we excluded data collecting during a washout period of 3 months following the ICD transition to minimize its effects. Finally, as these were surveillance data, our unit of analysis was ED encounter rather than ED patient.
In conclusion, in a statewide population of ED encounters with ADHF diagnoses, we did not observe significant sex differences in chief complaint categories. However, we did observe differences in chief complaints according to age. Comorbid conditions upon ED presentation may explain some of the age-related differences we observed, as older patients were more likely to present with cardiac dysrhythmias than younger patients. Characterizing the variation among presenting symptoms of ADHF patients in the ED may improve clinical care for patients with ADHF exacerbations.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Supplement1
Print 
