Lack of relationship between hypertension-associated symptoms and blood pressure in hypertensive ED patients

David J. Karras MD*, Jacob W. Ufberg MD, Richard A. Harrigan MD, David A. Wald DO, Maged S. Botros MD, Robert M. McNamara MD

Department of Emergency Medicine, Temple University School of Medicine, Philadelphia, PA 19140, USA

Received 16 December 2003; accepted 16 February 2004

Abstract A number of cardiopulmonary and neurological symptoms are presumed to be associated with hypertension. We examined the prevalence of these symptoms in ED patients with elevated blood pressure (BP) and studied the relationship between symptom prevalence and BP value. We enrolled consecutive adult ED patients with sustained BP elevation (systolic BP z140 mm Hg, diastolic BP z90 mm Hg). BP values were categorized according to Joint National Committee on Prevention, Evaluation, and Treatment of High Blood Pressure, 6th Report criteria. Elevated BP was noted in 551 (29%) of 1908 patients. Unprompted complaints of hypertension-associated symptoms were noted in 26%, and there was no association between BP category and complaints other than dyspnea. Symptom interviews were conducted in 294 (56%) patients; 68% of this subset noted z1 current hypertension-associated symptom with no relationship between symptom prevalence and BP category. We conclude that symptoms putatively associated with hypertension are common among ED patients with elevated BP, and their prevalence appears unrelated to BP value.

D 2005

Introduction

A number of cardiovascular, pulmonary, and neurological symptoms are believed to be associated with hypertension. Headache, visual changes, chest pain, dyspnea, and focal neurological deficits are widely considered potential symp- toms of blood pressure (BP) elevation and suggestive of acute hypertensive end-organ dysfunction [1-5]. Although not considered a sign of end-organ damage, epistaxis is also considered a hypertension-related symptom. The presence of

* Corresponding author. Tel.: +1 215 707 5032; fax: +1 215 707 3494.

E-mail address: david.karras@temple.edu (D.J. Karras).

any of these complaints in a patient with elevated BP may lead physicians to perform an extensive evaluation to exclude hypertensive emergency, defined as a situation requiring immediate BP reduction to mitigate acute end- organ dysfunction [2]. Some authorities regard the presence of these symptoms in a patient with significant BP elevation to be diagnostic of hypertensive emergency [1,6,7].

Many of the symptoms putatively associated with hypertension are common among patients presenting to the ED [8]. Although prior investigations have not detected a strong relationship between symptom prevalence and BP values among chronically hypertensive patients, these were studies of stable patient populations, and the findings may

0735-6757/$ - see front matter D 2005 doi:10.1016/j.ajem.2004.02.043

not be applicable to the ED population [9,10]. To our knowledge, there are no systematic studies in the ED or other acute care setting of the association between the degree of BP elevation and the presence of symptoms suggestive of hypertension-related end-organ damage.

We undertook a study of the association between the presence of hypertensive symptoms and BP values among ED patients with elevated BP. In part 1 of the study, chief complaints of all ED patients with elevated BP were extracted from the medical record. In part 2, patients were interviewed regarding current and recent symptoms. Our null hypothesis was that hypertension-associated symptoms, either as recorded in the medical record or elicited in interviews, are not more frequent among patients with greater degrees of BP elevation.

Methods

Study design

This was a prospective observational investigation. The protocol was approved by the medical school’s institutional review board. No change in medical management was required for this study. On-duty physicians, nurses, and students were not informed of the purpose of the investigation.

Study setting and population

The study was performed in the ED of an urban teaching hospital with approximately 50000 annual adult visits and serving a predominantly African-American population. The ED is staffed by attending physicians board-certified or board-eligible in emergency medicine, by emergency medicine residents, and by residents in other specialties performing emergency medicine rotations under the super- vision of the emergency medicine attending physicians.

Study protocol

Consecutive patients at least 18 years presenting to the ED during a 14-day enrollment period were eligible for participation. All patients had their BP measured by ED nurses using an automated BP device, calibrated to the manufacturer’s specifications (Welsh-Allyn Propaq Encore, Beaverton, Ore). The nurses had been trained by the manufacturer in the use of the device and were instructed to use a cuff of appropriate size. Patients arriving via triage had their BP measured in a seated position; those arriving by ambulance or taken directly to the treatment area had their BP measured while seated or supine on a stretcher. Patients with initial systolic BP values z140 mm Hg or initial diastolic BP values z90 mm Hg had their BP repeated at least 10 minutes later. Patients were included in the study if their systolic BP remained z140 mm Hg or diastolic BP remained z90 mm Hg upon repeat measurement. The only exclusion to enrollment was prior participation in the study.

Research associates were present in the ED at all times during the study period and were trained in data collection for this investigation. The associates reviewed all BP values recorded in triage and throughout the patients’ ED stay to determine study eligibility. Demographic information and all BP values obtained in the ED were recorded by the research associates on a standardized data collection instrument.

In part 1 of the study, the chief complaint of study patients was extracted from the medical record by the research associates. The primary source of this information was the triage nurses’ notes. For patients bypassing triage, the chief complaint was abstracted from the physician ED record. All patients meeting study enrollment criteria were included in part 1.

In the second part of the study, research associates approached all alert and oriented subjects for consent to undergo a detailed structured interview. Patients providing consent were questioned regarding the presence of symptoms typically cited as associated with hypertension, specifically, shortness of breath, chest pain, dizziness, headache, weak- ness, or numbness on one side of the body, nosebleed, and change in vision [1,6,7]. These symptoms are referred to hereafter as bhypertension-associated symptoms.Q Patients were asked whether each symptom was present at the time of ED arrival, in the past 24 hours, or in the week before ED presentation. Positive responses were not mutually exclusive.

Data analysis

Criteria established by the Joint National Committee on Prevention, Evaluation, and Treatment of High Blood Pressure, 6th Report (JNC-VI), were used to classify patients as having stage 1, stage 2, or stage 3 BP values, as presented in Table 1 [1]. When systolic and diastolic values were in different stages, patients were classified according to the greater stage. The greatest BP value obtained during the ED visit was used as the reference value in assigning BP stages for analysis.

Intergroup differences in the prevalence of each hyper- tension-associated symptom and cumulative symptom prevalence were assessed using the v² statistic or Fisher exact test, as appropriate, using a significance threshold of P b .05. Cochran’s linear trend statistic was used to assess intergroup trends in symptom prevalence. Independent assessments were planned of the data obtained in part 1, in which complaints were determined from the medical

Table 1 BP staging criteria
Category	Systolic BP (mm Hg)	Diastolic BP (mm Hg)
Not hypertensive	b140	b90
Stage 1	140-159	90-99
Stage 2	160-179	100-109
Stage 3	z180	z110
When systolic and diastolic BP values fall into different categories, the higher category is used to classify the subject. Adapted from the JNC-VI report [2].

Table 2 Prevalence of hypertension-associated symptoms noted in the chief complaint of ED patients with elevated BP (part 1)
Complaint	All	subjects (n = 529)	Stage 1 BP (n = 304)	Stage 2 BP (n = 127)	Stage 3 BP (n = 98)	P
Chest pain, n (%)	61	(12)	33 (11)	17 (13)	11 (11)
Shortness of breath, n (%)	58	(11)	25 (8)	13 (10)	20 (20)	.004
Headache, n (%)	27	(5)	18 (6)	3 (2)	6 (6)
Dizziness, n (%)	10	(2)	6 (2)	2 (2)	2 (2)
Nosebleed, n (%)	3	(1)	1 (0)	1 (1)	1 (1)
Blurred vision, n (%)	2	(0)	1 (0)	0 (0)	1 (1)
Focal neurological deficit, n (%)	5	(1)	2 (1)	2 (2)	1 (1)
Any complaint listed, n (%)	139	(26)	76 (25)	36 (28)	27 (28)
Data were obtained from the medical record; responses were not prompted by study staff. Unlisted P values are not significant (z.05).

record of all eligible patients, and the data obtained in part 2, in which symptoms were elicited from consenting patients.

Results

During 14 days of continuous data collection, 1908 patients presented to the ED. Elevated BP was noted in 551 (29%) patients. Twenty-two (4%) patients with elevated BP were excluded from further analysis because of incomplete data collection; none of these patients had consented to interview. Of the remaining 529 analyzable patients, 294 (56%) provided consent for interview, 200 (38%) were not approached for an interview due to severity of illness or inability to provide consent, and 35 (6%) declined to be interviewed.

The median subject age was 51 years (interquartile range [IQR], 39 to 65 years). Fifty-four percent of subjects were women. The majority of subjects (412 [78%]) were African American, with 43 (8%) Latino, 42 (8%) white (non- Latino), and 32 (6%) Asian subjects. The median systolic BP was 152 mm Hg (IQR, 142 to 170), and median diastolic BP was 88 mm Hg (IQR, 80 to 96). BP values were categorized as stage 1 in 304 subjects (57%), stage 2 in 127

(24%), and stage 3 in 98 (19%). No differences in sex, ethnicity, or BP values were noted between patients who consented for interview and those who did not. No differences were noted between excluded patients and those included in the final analysis.

Part 1 of the study assessed the prevalence of symptoms recorded in the medical record. As shown in Table 2, 26%

of study patients had chief complaints of one or more hypertension-associated symptoms. Chest pain and short- ness of breath were the most frequent symptoms. Dyspnea was found to be significantly more common among patients with stage 3 BP values than among patients with stage 1 or stage 2 BP values ( P = .004, Cochran’s linear trend P =

.002). No significant differences between the BP groups were noted in the prevalence of chest pain, dizziness, headache, focal neurological deficit, nosebleed, visual change, or headache complaints. No intergroup difference was noted in the prevalence of patients reporting at least one hypertension-associated symptom in their chief complaint.

In part 2, information regarding hypertension-associated symptoms was derived from patient interviews after prompting by research associates. Results are shown in Table 3. Sixty-eight percent of interviewed patients reported having one or more hypertension-associated symptoms at the time of arrival in the ED. No significant association between BP stage and the prevalence of any hypertension- associated symptom were noted. Further analyses revealed no significant intergroup differences in the prevalence of individual or aggregate hypertension-associated symptoms in the 24 h or 1 week before ED presentation.

Discussion

As many as 27% of adults in the United States are believed to have hypertension [11]. Reflecting the high prevalence of this condition, sustained BP elevation was noted in 29% of the adult ED patients in our study.

Table 3 Prevalence of symptoms at ED presentation among patients with elevated BP, obtained from structured interviews (part 2)
Symptom	All	subjects (n = 285)	Stage 1	BP (n = 145)	Stage 2	BP (n = 78)	Stage 3 BP (n = 62)
Chest pain, n (%)	74	(26)	41 (28)		21 (27)		12 (19)
Shortness of breath, n (%)	104	(36)	51 (35)		28 (36)		25 (40)
Headache, n (%)	92	(32)	50 (34)		23 (29)		19 (31)
Dizziness, n (%)	74	(26)	41 (28)		17 (22)		16 (26)
Nosebleed, n (%)	14	(5)	7 (5)		4 (5)		3 (5)
Blurred vision, n (%)	59	(21)	30 (21)		15 (19)		14 (23)
Focal neurological deficit, n (%)	3	(1)	2 (1)		0 (0)		1 (2)
Any symptom listed, n (%)	179	(63)	91 (63)		52 (67)		40 (64)
Patients were specifically asked by research associates if each listed symptom was currently present at the time of arrival. No intergroup differences are significant (all P z .05).

Although the majority of these patients have mildly elevated BP values, severely elevated BP is not uncommon in the ED. One large retrospective study in an urban ED documented stage 3 BP values in 2% of patients [12], whereas our prospective study found that 5% of adult ED patients met these criteria.

The assessment and management of patients with elevated BP have been extensively studied in primary-care environments. Evidence-based standards for classifying and treating hypertension have been established by the JNC [1]. In their sixth report (JNC-VI), the authors define hyperten- sion staging criteria that are widely cited in the medical literature and commonly used in assessing ED patients. The most recent JNC report (JNC-7) no longer employs this staging system [13].

The JNC authors specify that their hypertension manage- ment guidelines are intended for primary-care clinicians in assessing and managing BP elevation in patients without acute illnesses [2]. Extrapolation of the JNC guidelines to the ED setting may not be appropriate. ED patients often have acute illnesses or injuries that may elevate or depress their BP. It is generally not practical or possible to measure the BP according to JNC’s rigorous standards, and therapeutic decisions are often made without the repeated BP assess- ments advised in the JNC guidelines. Despite widespread adoption of the JNC-VI’s BP staging and management recommendations by emergency medicine textbooks and other guidelines for ED care [3,4], these guidelines were neither derived nor validated for use the ED setting [14].

The purpose of this study was to determine if symptoms putatively regarded as associated with hypertension are more prevalent among ED patients with greater degrees of BP elevation. Failure to detect a relationship between symptom prevalence and BP value would bring into question the validity of ascribing these symptoms to BP elevation. An unprompted complaint of dyspnea was, in fact, found to be more common among patients with greater degrees of BP elevation. However, no other complaints potentially suggestive of acute cardiovascular, cerebral, or visual dysfunction differed significantly between patients in any BP group.

Patient interviews revealed that symptoms often attrib- uted to hypertension are highly prevalent among ED patients with elevated BP. When patients with elevated BP were prompted for symptom presence, more than two thirds reported having at least one symptom potentially attribut- able to hypertension. No single symptom was more frequent among patients with greater degrees of BP elevation, and the overall prevalence of any hypertension-related symptom did not differ between BP groups.

Hypertensive emergencies are defined as relatively rare situations requiring immediate reduction in BP to prevent or limit acute damage to the brain, heart, eyes, and kidneys [1-7,15]. Many authorities suggest that in the setting of significantly elevated BP, hypertension-associated symp- toms should be presumed to reflect acute end-organ injury

and that the presence of these symptoms distinguishes hypertensive emergency from a non-critical hypertensive urgency [1,6,7]. These criteria may be overly simplistic and would lead to the diagnosis of hypertensive emergency in the majority of ED patients with elevated BP in our study. Our findings are consistent with an Italian study noting hypertension-associated symptoms in 28% of ED patients with severe BP elevation [16]. The authors of that study concluded that hypertensive emergencies are common events in the ED population. An alternative explanation, however, is that hypertension-related symptoms are so nonspecific as to make their presence alone inadequate to diagnose hypertensive emergency.

Limitations

We studied only ED patients with elevated BP. Stronger conclusions regarding the association of hypertension- associated symptoms and BP magnitude require that symptom prevalence be determined in ED patients with normal BP values. It is possible that some study patients would be normotensive under other circumstances or that BP values would normalize if further measurements were taken in the ED [17].

As previously mentioned, the BP measurement standards used in the JNC reports were not intended for the ED and are impossible to strictly apply in this setting. We did not train the ED staff in the BP measurement techniques advised in the report. Despite the flaws inherent in applying the JNC BP staging criteria to ED patients, these criteria are widely used in this setting, frequently cited in the emergency medicine literature, and familiar to emergency physicians. We believe our BP measurement practices are typical of those in other EDs.

This study was conducted in a single, urban, academic ED, and the majority of our subjects were African American. Our findings may not be generalizable to other ED settings or other populations. Larger multicenter studies are needed to confirm these results.

Recall bias may play a significant role in our calculation of symptom prevalence. It is possible that prompting by the research associates in part 2 of the study may have caused patients to overstate the prevalence of hypertension- associated symptoms. This bias is unlikely to have effected the prevalence of hypertension-associated symptoms in part 1 which were derived from the medical record. Because we relied on patient-reported complaints rather than physician diagnoses, it is possible that some complaints, such as chest pain, were not cardiovascular in etiology and that some patients would not be classified as having symptomatic hypertension.

Our failure to detect significant intergroup differences in symptom prevalence may have been caused by type II error. Epistaxis was the chief complaint of only 0.3% of subjects, and retrospective power calculations show that more than 7000 subjects in each BP group would have been necessary

to detect a doubling in prevalence of this complaint. A recent prospective study has, in fact, demonstrated active epistaxis to be more common among patients with greater BP values [18]. For the more common symptoms noted in part I of our study, such as dyspnea, chest pain, and headache, our sample size was sufficient to detect an absolute intergroup difference in symptom prevalence of about 15% with a power of 80%. Other than the prevalence of dyspnea, we noted no trends toward intergroup differ- ences in symptom prevalence.

Conclusions

We conclude that the symptoms commonly regarded as being associated with hypertension are highly prevalent among ED patients with elevated BP. The majority of patients with elevated BP values report having at least one such symptom, and there is no relationship between the prevalence of most symptoms and the patient’s BP stage. Physicians should exercise caution when attributing symp- toms to BP elevation in this population.

Acknowledgment

The authors thank Jane Prosser, MD, for valuable assistance with data analysis and entry; to Kathleen Hatala, BSN, for assistance with data management; and to Christine Shields, RN, for supervising data acquisition.

References

1. Calhoun DA, Oparil S. Treatment of hypertensive crisis. N Engl J Med 1990;323:1177 - 83.
2. Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure. The sixth report of the Joint

National Committee on Prevention, Detection, Evaluation and Treat- ment of High Blood Pressure. Arch Intern Med 1997;157:2413 - 45.

1. Mathews J. Hypertension. In: Rosen P, Barkin R, editors. Emergency medicine: concepts and clinical practice. 4th ed. St Louis: Mosby-Year Book Inc; 1998. p. 1755 - 70.
2. Wu MM, Chanmugam A. Hypertension. In: Tintinalli JE, Kelen GD, Stapczynski JS, editors. Emergency medicine: a comprehensive study guide. 5th ed. New York7 McGraw-Hill; 2000. p. 401 - 12.
3. Gifford RW. Management of hypertensive crises. JAMA 1991; 266:829 - 35.
4. Ram CVS. Immediate management of severe hypertension. Cardiol Clin 1995;13:579 - 91.
5. Kitiyakara C, Guzman NJ. Malignant hypertension and hypertensive emergencies. J Am Soc Nephrol 1998;9:133 - 42.
6. McCaig L, Ly N. National hospital ambulatory medical care survey: 2000 emergency department summary. Advance data from vital and health statistics; no. 326. Hyattsville (Md)7 National Center for Health Statistics; 2002.
7. Weiss NS. Relationship of high blood pressure to headache, epistaxis, and selected other symptoms. N Engl J Med 1972;287:631 - 3.
8. Kottke TE, Tuomilehto J, Puska P, Salonen JT. The relationship of symptoms and blood pressure in a population sample. Int J Epidemiol 1979;8:355 - 9.
9. Hyman DJ, Pavlik VN. Characteristics of patients with uncontrolled hypertension in the United States. N Engl J Med 2001;345:479 - 86.
10. Chiang WK, Jamshahi B. Asymptomatic hypertension in the emergency department. Am J Emerg Med 1998;16:701 - 4.
11. Chobanian AV, Bakris GL, Black HR, et al. Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA 2003; 289:2560 - 72.
12. Shayne PH, Pitts SR. Severely increased blood pressure in the emergency department. Ann Emerg Med 2003;41:513 - 29.
13. Varon J, Marik PE. The diagnosis and management of hypertensive crisis. Chest 2000;118:214 - 27.
14. Zampaglione B, Pascale C, Marchisio M, Cavallo-Perin P. Hyperten- sive urgencies and emergencies: prevalence and clinical presentation. Hypertension 1996;27:144 - 7.
15. Pitts SR, Adams RP. Emergency department hypertension and regression to the mean. Ann Emerg Med 1998;31:214 - 8.
16. Herkner H, Laggner AN, Mullner M, et al. Hypertension in patients presenting with epistaxis. Ann Emerg Med 2000;35:126 - 30.