Prevalence of Neurological Complaints among Emergency Department Patients with Severe Hypertension

Abstract

Objective

Severe hypertension can accompany neurological symptoms without obvious signs of target organ damage. However, acute cerebrovascular events can also be a cause and consequence of severe hypertension. We therefore use US population-level data to determine prevalence and clinical characteristics of patients with severe hypertension and neurological complaints.

Methods

We used nationally representative data from the National Hospital Ambulatory Medical Care Survey (NHAMCS) collected in 2016–2019 to identify adult ED patients with severely elevated blood pressure (BP) defined as systolic BP ≥ 180 mmHg and/or diastolic BP ≥120 mmHg. We used ED reason for visit data fields to define neurological complaints and used diagnosis data fields to define acute target organ damage. We applied survey visit weights to obtain national estimates.

Results

Based on 5,083 observations, an estimated 40.4 million patients (95% CI: 37.5–43.0 million) in EDs nationwide from 2016–2019 had severe hypertension, equating to 6.1% (95% CI: 5.7–6.5%) of all ED visits. Only 2.8% (95% CI: 2.0–3.9%) of ED patients with severe hypertension were diagnosed with acute cerebrovascular disease; hypertensive urgency was diagnosed in 92.0% (95% CI: 90.3–93.4%). Neurological complaints were frequent in both patients with (75.6%) and without (19.9%) cerebrovascular diagnoses. Hypertensive urgency patients with neurological complaints were more often older, female, had prior stroke/TIA, and had neuroimaging than patients without these complaints. Non-migraine headache and vertigo were the most common neurological complaints recorded.

Conclusion

In a nationally representative survey, one-in-sixteen ED patients had severely elevated BP and one-fifth of those patients had neurological complaints.

Introduction

Hypertension is a leading cause of cardiovascular disease and premature death worldwide as well as a major risk factor for stroke.^1,2 The prevalence of hypertension is substantial with nearly half of US adults (47% or 116 million people) diagnosed with hypertension based on measured blood pressure (BP) values or prescription of anti-hypertensive medications.^1,3 Unsurprisingly, elevated BP is common among emergency department (ED) patients and the proportion of ED visits for hypertensive emergencies has increased monotonically in the US.^4,5

The brain is one potential organ that can be damaged by acute severe hypertension.⁶ Additionally, acute cerebrovascular events are themselves associated with BP elevations.^7,8 Diagnosing acute vascular brain injury either as a consequence of or a reason for severe hypertension can be challenging given the variety of potential neurological manifestations of cerebrovascular disease ranging from encephalopathy to minor, transient, and non-specific neurological deficits.^9–11

Our work therefore aims to determine the prevalence and types of neurological complaints reported among US ED patients with severe hypertension using population-level data. To further understand current care practices for these ED patients with severely elevated BP, we also detail their demographic, clinical, management, and disposition characteristics comparing those presenting with versus without neurological complaints.

Methods

Study Design

We used National Hospital Ambulatory Medical Care Survey (NHAMCS) data from 2016 through 2019 to identify adult ED patients with severely elevated BP defined as an initial ED reading of ≥ 180 mmHg systolic and/or ≥120 mmHg diastolic.^12,13 The NHAMCS collects information from a nationally representative random sample of EDs of nonfederal hospitals using a 4-stage random sampling design that narrows from geographic areas to individual hospitals to particular departments and then to a set of patient visits annually. In the selected EDs, staff undergo training on survey methodology and then use structured forms to collect information about a random sample of patient visits over a 4-week period. These data reflect 87% to 98% participation by the approximately 10% of US EDs surveyed each year. All NHAMCS data, including coding of free-text fields, undergoes multiple consistency checks at the National Center for Health Statistics.¹⁴ Our study was exempt from review by our institutional review board since all NHAMCS data is publicly available and deidentified.

Participants

We included all ED patients aged >18 years with severely elevated BP as measured on their initial vital signs recording. We subclassified patients by whether there was evidence of acute target organ damage (hypertensive emergency) or not (hypertensive urgency).¹² We defined evidence of target organ damage as having a probable, questionable, or rule-out ED diagnosis of: (1) myocardial infarction, (2) heart failure, (3) acute renal failure, (4) aortic dissection, (5) eclampsia, and (6) cerebrovascular disease (ischemic stroke, hemorrhagic stroke, subarachnoid hemorrhage, hypertensive encephalopathy, posterior reversible encephalopathy/reversible cerebral vasoconstrictive syndrome, or TIA). We used validated International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes in any position to define these conditions. Specifically we used: I21.x, I22.x, I23.x for myocardial infarction;¹⁵ I50.x for heart failure;¹⁶ N17.x for acute renal failure;¹⁷ I71.x for aortic dissection;¹⁸ O15.x for eclampsia;¹⁹ and I60.x, I61.x, I63.x, I64.x, I67.4, I67.841, I67.83, G45.x, or H34.1 for cerebrovascular disease.^20,21 In NHAMCS, up to five diagnoses are coded in sequence from free text responses by ED staff which are then translated into ICD-10-CM codes by medical billers based on coding instructions in the NHAMCS Coding Requirements Manual.¹⁴ For all of the diagnoses recorded, the NHAMCS collection instrument allows staff to indicate if the diagnosis is probable, questionable, or rule-out.

Measurements

We identified patients with versus without neurological complaints using coded reason for visit information. In the NHAMCS instrument, trained study staff are instructed to “list the first 5 reasons for visit (i.e., symptoms, problems, issues, concerns of the patient) in the order in which they appear. Start with the chief complaint and then move to the patient history for additional reasons.” Free text information entered into the ‘Reason for Visit’ field is then coded using a previously developed modular classification system and added to the NHAMCS dataset. The module classification system was originally developed iteratively by a panel of healthcare professionals to assure its clarity, accuracy, specificity, efficiency, consistency, clarity, flexibility of revision, and computer applicability; the system is updated as needed to incorporate new codes as well as change existing codes.²² For the current study, we defined patients with neurological complaints as those in whom any of the 5 listed reasons for visit included a neurological symptom, sign, condition, or problem based on the modular classification system; all complaints included in the “symptoms referable to the nervous system” module were considered neurological (Table 1).

Table 1.

Neurological Complaints based on Reason for Visit

Reason for visit	Entities include
Abnormal involuntary movements	Jerking Shaking Tics Tremors Twitch
Altered level of consciousness	Altered level of consciousness
Cerebrovascular disease	Berry / brain aneurysm (added in 2019) Cardiac aneurysm (added in 2019) Carotid stenosis Cerebral aneurysm NOS (added in 2019) Cerebral arteriosclerosis Cerebral hemorrhage Cerebral stenosis Cerebrovascular accident (CVA) Stroke TIA
Convulsions	Febrile convulsions Fits Seizures Seizure Disorders Spells
Difficulty in swallowing (dysphagia)	Choking Inability to swallow
Disturbances of memory	Amnesia Forgetfulness Lack or loss of memory Temporary loss of memory
Disorders of motor function	Ataxia Clumsiness Difficulty in walking Dyskinesia Fall, NOS Limping Staggering Stumbling Uncoordinated Unsteady gait
Disturbances of sensation	Lack of Sensation No response to pain Numbness Stunned Increased sensation Burning legs Burning, tingling sensation Needles and pins Prickly feeling Stinging Other disturbances of sense, including smell and taste
Disorders of speech, speech disturbance	Groping for words Speech/verbal apraxia Unable to speak Stuttering, stammering Slurring
General weakness	General weakness
Headache	Post-traumatic headache Sinus headache Tension headache Symptoms of head, not elsewhere classified
Migraine headache	Migraine headache
Other symptoms referable to the nervous system	Confusion Cognitive decline Damaged nerves Neuralgia Neurovegetative Pinched nerve Post-ictal Apraxia Speech/verbal apraxia
State of consciousness not specified	Found on floor Verbally unresponsive
Unconscious on arrival	Coma Found unconscious Knocked out Shock Stupor Unresponsive
Vision Dysfunctions	Visual field deficit Blurred vision Can’t see distances Difficulty reading Eyes won’t focus Obstruction, vision Poor vision Scotoma Trouble seeing Weak eyes Cloudy vision Fuzzy vision Spots in front of eyes (floaters) Diplopia
Vertigo - dizziness	Falling sensation Giddiness (dizziness) Lightheadedness Loss of sense of equilibrium or balance Room spinning
Voice Disorders	Hoarseness, loss of voice Croaky Hypernasality
Weakness (neurological)	Drooping, facial or not otherwise specified Right- or left-sided weakness

To describe identified patients with severely elevated BP, we noted patient age, sex, race-ethnicity (white non-Hispanic, black non-Hispanic, Hispanic, or other, defined as Asian, Native Hawaiian/other Pacific Islander, American Indian/Alaska Native, or of more than 1 race), and payment type (private, Medicare, Medicaid, or other insurance). To describe the clinical characteristics of our cohort, we recorded the following vascular comorbidities which were explicitly asked about in the NHAMCS instrument: hypertension, diabetes, coronary artery disease, congestive heart failure, hyperlipidemia, chronic kidney disease, cancer, prior stroke/TIA, substance abuse, and alcohol abuse.^23,24 To evaluate management, we included diagnostic testing performed in the ED (CT and/or MRI), if patient was seen by any consulting physician, whether or not cardiovascular (e.g., anti-hypertensives, diuretics, vasodilators) and/or coagulation modifying (e.g., anticoagulants, antiplatelet agents, anticoagulant reversal agents) agents were administered at index ED visit or prescribed at the time of ED discharge, and ED disposition (left against medical advice or before being seen, transferred to another facility [nursing home, psychiatric hospital, or other], discharged to home, admitted to an observation unit, and hospitalized). To assess regional patterns, EDs were categorized into geographic regions defined by NHAMCS: Northeast, Midwest, South, and West.

Analysis

We applied survey visit weights to obtain national estimates. We report 95% confidence intervals (CI) to measure the degree of uncertainty in the NHAMCS sampling method. We use a survey-weighted student t-test to compare age and survey-weighted chi-squared testing to compare all other demographic, clinic, management, disposition, and regional variables between hypertensive urgency patients with neurological complaints versus those without neurological complaints. No adult patients included in NHAMCS with severely elevated BP were excluded from our study, however, estimates based on less than 30 unweighted records are not reported in keeping with NHAMCS guidelines.²⁵ The overall threshold of statistical significance was set at an α = 0.05. Analyses were performed with Stata MP (version 14.2; StataCorp LP, College Station, TX) using the svy command to account for the survey methodology.

Data Availability

The NHAMCS data that support the findings of this study can be obtained directly from the Centers for Disease Control and Prevention. The analytical code we used will be made available upon reasonable request from the corresponding author.

Results

Overall Population Characteristics

Based on 5,083 observations, an estimated 40.4 million patients (95% CI: 37.5–43.0 million) were evaluated in EDs nationwide from 2016 through 2019 with severely elevated BP, equating to 6.1% (95% CI: 5.7–6.5%) of all ED visits. Among these patients, based on 1,089 observations, an estimated 8.6 million (95% CI: 7.7–9.6 million) had neurological complaints equating to 21.4% (95% CI: 19.2–23.9%) of all visits with severe hypertension. Based on 743 observations, patients with neurological complaints other than headache accounted for 15.5% (95% CI: 13.3–17.9%) of all ED visits with severe hypertension.

Patients with Target Organ Damage

Based on 388 observations, an estimated 3.2 million patients (95% CI: 2.6–3.9 million) with severe hypertension had target organ damage diagnosed (hypertensive emergency), equating to 8.0% (95% CI: 6.6–9.7%) of patients with severe hypertension (Figure 1). Among these, an estimated 1.1 million patients (95% CI: 0.8 –1.6 million) were diagnosed with cerebrovascular disease, representing 35.2% (95% CI: 25.8–45.9%) of all patients with hypertensive emergency and 2.8% (95% CI: 2.0–3.9%) of all ED patients with severely elevated BP. In the subgroup of patients with hypertensive emergency, neurological complaints were reported in 21.6% of patients without cerebrovascular disease. Overall, neurological complaints were more common in patients with cerebrovascular disease diagnosed than others, but neurological complaints were nevertheless also common in patients with severely elevated BP without a diagnosis of cerebrovascular disease (75.6% vs. 19.9%).

Figure 1.

Study flow

Patients without Target Organ Damage

Based on 4,695 observations, an estimated 37.1 million patients (95% CI: 34.4–39.6 million) in EDs nationwide from 2016–2019 had severely elevated BP without target organ damage (hypertensive urgency), equating to 5.6% (95% CI: 5.2–6.0%) of all ED visits. ED visits with hypertensive urgency accounted for 92.0% (95% CI: 90.3–93.4%) of visits with severe BP elevation. Based on 946 observations, an estimated 7.3 million patients (95% CI: 6.5–8.1 million) with hypertensive urgency in EDs nationwide from 2016–2019 had neurological complaints, equating to 19.7% (95% CI: 17.7–22.0%) of patients with hypertensive urgency.

In the subgroup of hypertensive urgency patients with neurological complaints, the most frequent neurological symptom was non-migraine headache in 36% of patients followed by vertigo/dizziness in 26% (Table 2).

Table 2.

Neurological Complaints Reported among Patients with Hypertensive Urgency (n=946)

Specific neurological complaints	Percentage (95% CI) of hypertensive urgency visits with neurological complaints
Headache	36.39 (30.50–42.63)
Vertigo/dizziness	25.74 (20.70–31.52)
General Weakness	13.67 (10.38–17.78)
Disturbances of sensation	12.33(8.73–17.12)
Disorders of motor functions	7.57 (5.02–11.25)
Other symptoms referrable to the nervous system	4.67 (3.11–6.95)
Abnormal involuntary movements	4.30 (2.49–7.34)
Convulsions	3.88 (2.12–6.99)
Unconscious on arrival	2.60 (1.50–4.49)
Migraine headache	1.49 (0.77–2.87)
Vision dysfunctions	3.56 (2.33–5.42)
Dysphagia	2.32 (1.06–5.02)
Cerebrovascular disease	1.68 (1.03–2.73)
Neurological weakness	1.15 (0.53–2.49)
Speech disturbances	0.97 (0.24–3.77)
State of consciousness issue	0.74 (0.27–2.0)

The mean age of patients with ED visits with hypertensive urgency was 61 years. These patients were 61.3% female and 62.1% white. A medical history of hypertension was reported in 70.8% patients. The majority of ED visits (76.8%) with hypertensive urgency resulted in discharge to home (Table 3).

Table 3.

Characteristics of Patients with Hypertensive Urgency

	Percentage of national ED visits with hypertensive urgency
	Total	Any neurological symptom listed as reason for visit	No neurological symptom listed as reason for visit	P-value
Demographics
Mean Age	61.04	63.02	60.55	0.04
Sex
Male	38.67	31.65	40.38	0.01
Female	61.34	68.35	59.62
Race-ethnicity
White, non-Hispanic	62.13	63.21	61.86	0.85
Black, non-Hispanic	24.91	23.34	25.30
Hispanic	10.18	10.88	10.01
Other	2.77	2.57	2.82
Payment Source
Private insurance	35.99	40.21	34.95	0.52
Medicare	41.81	41.69	41.84
Medicaid	24.15	25.63	23.79
Other	23.76	20.21	24.63
Region
Northeast	12.73	11.62	13.00	0.42
Midwest	22.55	25.91	21.73
South	42.97	40.80	43.50
West	21.75	21.67	21.77
Medical History
Hypertension	70.78	74.16	69.95	0.18
Diabetes	27.13	29.44	26.56	0.34
Congestive heart failure	5.24	4.76	5.36	0.58
Coronary artery disease	12.56	12.79	12.50	0.88
Cancer	5.67	4.93	5.85	0.52
Stroke/TIA	6.53	11.03	5.42	<0.001
Chronic kidney disease or end stage renal disease	8.18	7.31	8.38	0.53
Substance Abuse	4.82	3.82	5.06	0.28
Alcohol Abuse	2.56	3.90	2.24	0.13
Management
Head CT obtained in ED	13.53	35.67	8.09	<0.0001
Any MRI obtained	1.24	2.42	0.95	0.04
Seen by consulting physician	13.03	18.40	11.71	0.002
Treatment with at least one cardiovascular agent	31.75	37.32	30.38	0.02
Treatment with at least one coagulation modifying agent	4.07	6.79	3.40	<0.001
ED disposition
Left against medical advice or before being seen	1.39	1.17	1.44	0.001
Transferred to another facility (Nursing home, psychiatric facility, other)	3.76	7.24	2.90
Discharged to home	76.78	71.33	78.12
Admitted to an observation unit	2.48	2.56	2.47
Hospitalized	13.36	17.77	12.27

Comparing hypertensive urgency patients with versus without neurological complaints, those with neurological complaints were older with mean age (63.0 vs. 60.6 years; P=0.04) and more often female (68.4% vs. 59.6%; P=0.01). There were no differences between groups in terms of race-ethnicity or payment source. Hypertensive urgency patients with neurological complaints were more likely to have prior stroke/TIA (11.0% vs. 5.4%; P<0.001) as compared to those without neurological complaints; past medical history was otherwise similar. There were no significant regional differences between patients with hypertensive urgency with versus without neurological complaints (Table 3).

In the ED, head CT was obtained far more often among hypertensive urgency patients with neurological complaints than those without these complaints (35.7% vs. 8.1%; P<0.0001). Although MRI was infrequently obtained overall, hypertensive urgency patients with neurological complaints also had this test performed more often than those without neurological complaints (2.4% vs. 1.0%; P=0.04). Hypertensive urgency patients with neurological complaints were also more likely to be seen by a consultant in the ED (18.4% vs. 11.7%; P=0.002), treated with at least one cardiovascular agent (37.3% vs. 30.4%; P=0.02) as well as at least one coagulation modifying agent (6.8% vs. 3.4%; P<0.001). In terms of ED disposition, patients with neurological complaints were more likely to be hospitalized (17.8% vs. 13.4%) or transferred to another facility (7.2% vs. 3.8%) than other hypertensive urgency patients (Table 3).

Discussion

Using data from NHAMCS, we found that severely elevated BP was initially recorded in 6% of all US ED visits from 2016–2019. Among patients with hypertensive urgency, which accounted for the vast majority of visits with severe hypertension, nearly one-fifth had a neurological complaint, corresponding to approximately 1.8 million patient visits per year.

Prior data regarding the proportion of ED patients with severely elevated BP without diagnosed target organ damage is based primarily on single center, retrospective studies. A recent meta-analysis designed to better understand the ED prevalence of hypertensive crises included 8 studies all conducted in countries other than the US. These researchers found that the prevalence of hypertensive emergency and urgency visits were 0.3% and 0.9% of all ED visits, respectively. But this estimate showed high variability across the different studies with the prevalence of hypertensive urgency ranging between 0.24% and 2.4% (I² 100%).⁹ Using NHAMCS, we found a much higher prevalence of ED visits with severely elevated blood pressure as compared to this meta-analysis as well as to a US study of ED visits for hypertensive emergency using administrative claims data from 2006–2013.⁵ Our higher prevalence likely reflects our exclusive use of initial ED vitals to define the cohort. Indeed, a recent Korean study found that 6% of all ED patients at a single regional medical center had an initial systolic BP ≥180 mmHg and/or diastolic BP ≥100 mmHg.²⁶ Alternatively, our data may reflect the recent worsening of blood pressure control among patients with hypertension nationally.²⁷

Our finding that one-fifth of ED patients with hypertensive urgency have neurological complaints is in keeping with prior literature. In a prior study of 4,983 ED patients with hypertensive urgency, 22% had headache complaints and 7% had other neurological symptoms.⁹ Detailed descriptions of symptoms among patients with hypertensive urgency from single and multicenter retrospective studies include complaints of focal neurological deficits, headache, vertigo/dizziness, confusion, and paresthesias.^28–30 Our results regarding the specific types of neurological complaints among patients with severe hypertension are also similar to those reported previously with non-migraine headache as most common.^5,9,31 Our work adds to extant literature by providing US population level estimates of the prevalence and types of neurological complaints experienced by ED patients with elevated BP. Given the proportion of patients with non-specific neurological complaints, including headache, that we identified, our work highlights the importance of continued research to develop risk-stratification tools and evidence-based treatment protocols in acute diagnostic neurology which is a field explicitly concerned with the initial assessment, diagnosis, management, and referral of patients presenting with new neurologic symptoms that are not obviously due to serious, life-threatening neurologic diseases…but might be.³²

Overall, ED resource utilization for hypertensive urgency patients with neurological complaints, as reflected by frequency of consultation, imaging obtained, and relevant medical treatments given, was higher compared to other hypertensive patients. While a higher rate of head CT and MRI among ED patients with neurological complaints as compared to those without these complaints is to be expected, the fact that close to a third of patients with neurological complaints in our study received a head CT may suggest a potential subgroup of ED patients to target for improvement in the diagnosis of cerebrovascular disease. Among ED patients with non-specific neurological discharge diagnoses, including primary headache conditions and vertigo, those who have a head CT obtained prior to ED discharge home have increased risk of subsequent stroke hospitalization.^33–35 This might represent selective imaging of higher-risk older patients who later on develop a stroke or that, in a small subset of patients, concurrent strokes present at index ED visit were missed because small and hyperacute infarcts are not well seen on head CT.^35,36 Future research with longitudinal patient follow up is needed to determine which ED patients amongst those with negative head CT imaging who present with neurological complaints need closer outpatient follow up and/or additional neuroimaging.

Our study has several important limitations. To begin with, different BP cut offs have been used in the literature as well as by guidelines to define severely elevated BP and hypertensive crises.^12,13,37 Our choice of including patients with BP ≥180/120 mmHg, though based on an established definition, is thus somewhat arbitrary but represents an advantage over the use of administrative claims data alone to identify patients with severe hypertension given variable definitions and inconsistent real-world use of the terms malignant hypertension, hypertensive emergency, hypertensive urgency, and asymptomatic hypertension.¹³ However, our exclusive use of initial ED measurement values to define our study cohort, rather than only patients labeled as having a hypertensive urgency or emergency by treating providers or using re-measured BP values may be overly inclusive and differs from definitions used in prior studies.^5,9,38 Failure to consider blood pressure measurements taken after the initial one in the ED may have lead us to overestimate rates of severe hypertension. Second, since there is no well-established definition of neurological complaints in the literature, different researchers may have categorized different reasons for visit as neurological. Similarly, since headache are difficult to characterize in the ED setting,³⁹ patients with a number of different primary headache disorders may have been miscategorized as having or not having a migraine headache as their reason for visit. Third, our use of ED probable, questionable, and rule out diagnoses to classify acute target organ damage is imperfect. It is possible that patients’ diagnoses changed during their ED work up or, for the small percentage of patients in our cohort who were hospitalized, changed during their inpatient stay. Fourth, in NHAMCS, there is no field for MRI brain and therefore our ascertainment of MRI studies could have included imaging of non-brain regions. Finally, we are unable to report the frequency of specific subspecialty (e.g., neurology, neurosurgery) consultations as well as whether included patients had a history of atrial fibrillation at the time of ED arrival due to the absence of this information in NHAMCS.

Conclusion

We identified a high volume of US ED patients with severely elevated BP as measured at initial ED presentation. About 8% of these patients had hypertensive emergency, and among the remainder, nearly 20% had neurological complaints such as headache and vertigo. Given the nonspecific nature of these neurological complaints, and the close association between elevated BP and acute cerebrovascular events, further research is required to define efficient management strategies as well as tools to identify the presumably small subset of patients with severely elevated BP and neurological complaints who are at increased short-term stroke risk.