Elevated blood pressure during emergency departments visit is associated with increased rate of hospitalization for heart failure: A retrospective cohort study

Abstract

There is a limited literature discussing the long‐term outcome of patients admitted to the emergency department with elevated blood pressure. The aim of the present study was to evaluate outcomes of patients with hypertension who attended an emergency department. All patients with hypertension who attended an emergency department without target organ damage were evaluated. A composite end point at 18 months, which included all‐cause mortality, acute coronary syndrome, cerebrovascular accident, or hospitalization for heart failure, were compared between patients with hypertension and those with normotension. Overall, 410 patients were included in the study. Baseline characteristics were similar between patients with hypertension and those with normotension, except chronic renal failure being more prevalent in patients with hypertension. The composite primary end point occurred similarly in both groups; however, hospitalization for heart failure was significantly more common in patients with hypertension. Elevated blood pressure during an emergency department visit is associated with an increased risk for hospitalization for heart failure during an 18‐month follow‐up period compared with normotension.

1. INTRODUCTION

Approximately 1% to 2% of patients who constantly receive antihypertensive drugs will undergo acute and severe elevation of blood pressure (BP) throughout their lifetime.1, 2, 3, 4 Over time, hypertension‐related emergency department (ED) visits have increased, although an increasing number of patients are discharged from the ED.5 This means that most patients attending the ED with elevated BP are followed up only in ambulatory settings without observation in the hospital. Thus, it is important to identify patients at risk for adverse outcomes based on ED data. It is well known that early diagnosis and management of hypertension reduce long‐term morbidity and mortality, predominantly vascular risk, in outpatient settings6, 7; however, evidence regarding the long‐term clinical outcome of hypertensive urgency is currently lacking in the literature. Although there are clear guidelines for initiation of antihypertensive therapy in the outpatient setting, such guidelines do not exist for treatment of elevated BP in the ED2, 5, 8 and there is no consensus regarding the necessity of lowering BP in patients with hypertensive urgency. A recent randomized controlled trial comparing the clinical efficacy of resting and antihypertensive medication in patients with hypertensive urgency found that there was no significant difference between the two groups.9 This finding suggests that perhaps resting in the management of hypertensive urgency has clinical efficacy. Several studies were conducted in an attempt to evaluate the optimal treatment approach for hypertensive urgency10, 11, 12, 13, 14; however, only a few included long‐term follow‐up of the patients. Merlo4 and Vlcek15 and colleagues have shown an association between hypertensive urgencies and subsequent cardiovascular events in follow‐up periods of 12 and 24 months, respectively. Both of these studies evaluated the clinical significance of hypertensive urgency, but the significance of elevated BP, which is not in the range defining hypertensive urgency, is ill‐defined. In addition, according to a large study by Niiranen and colleagues,16 isolated diastolic and systolic hypertension defined with home measurements, but not office‐measured BP, were found to be associated with increased cardiovascular risk. A recently published article by Patel and colleagues17 aimed to describe the short‐term outcomes of patients with hypertensive urgency and compare outcomes of outpatient and hospital management found that although hypertensive urgency is common, the rate of major adverse cardiovascular events in asymptomatic patients was low. Furthermore, they found that visits to the ED were associated with more hospitalizations but not improved outcomes. The aim of the present study was to evaluate the long‐term outcomes of patients attending the ED with elevated BP.

2. METHODS

2.1. Study population

Our study was a retrospective cohort trial performed from 2012 to 2015 in the Rabin Medical Center ED, which is the largest ED in Israel, with over 150 000 visits annually. All patients aged 18 to 100 years who were discharged from the ED with a primary diagnosis of hypertension and were asymptomatic or had symptoms of headache, chest pain, dyspnea, weakness, gastrointestinal symptoms, recurrent falls, blurred vision, bleeding, or other unspecified symptoms, were included. Elevated BP in the ED was defined as a systolic BP ≥140 mm Hg or diastolic BP ≥90 mm Hg. Patients were excluded if they were admitted to the ED with target organ damage such as acute myocardial infarction/acute coronary syndrome (ACS), pulmonary edema, cerebrovascular accident/transient ischemic attack, acute renal failure, encephalopathy, or a clear infectious etiology. Demographic variables, BP and heart rate, relevant laboratory values, clinical manifestations at the ED visit, treatment in the ED, comorbidities, and previous hypertensive medications were extracted from the patients’ computerized medical records. Patients were followed up for 18 months after the ED visit for the occurrence of the primary outcome. Primary outcome was the occurrence of either ACS/nonfatal myocardial infarction, CVA/transient ischemic attack, hospitalization for congestive heart failure (CHF)/left ventricular failure, atrial fibrillation, and all‐cause mortality in addition to a composite end point, including only all‐cause mortality, ACS, CVA, or hospitalization for CHF. Furthermore, to examine the influence of hypertensive events on renal function, the difference between the creatinine level after the 18‐month follow‐up and the creatinine level on admission to the ED was calculated. BP medications dispensed following discharge and compliance were also recorded. Compliance was defined based on the percentage of antihypertensive medications purchased during the follow‐up period. Information was retrieved from the patients’ medical records and, if unavailable, through telephone calls to the treating physician or to the patient. At the end of the follow‐up period, two different and independent researchers extracted the data regarding patient outcomes from the medical files and the information was crosschecked to ensure the validation of the data. Clinical outcomes of patients admitting to the ED with elevated BP were compared with age‐, sex‐, and history of hypertension–matched groups of patients attending the ED for a random cause. We manually performed 1:1 matching by age, sex, and previous diagnosis of hypertension before the visit to the ED. The study was approved by the Rabin Medical Center institutional review board.

2.2. BP measurement and data collection

BP measurement was performed by a nurse or a physician with the patient in a sitting or supine position following at least 5 minutes of rest. When systolic BP was ≥140 mm Hg or diastolic BP ≥90 mm Hg, a second BP measurement was performed to confirm the values at least 5 minutes following the initial measurement. BP was measured in a single arm, which was arbitrarily chosen by the nurse or physician, unless the patient requested the BP to be measured in a certain arm because of a history of higher BP values in that arm. BP measurements were peformed with a standard sphygmomanometer (Vital Signs Monitor 52 NTP model, Welch Allyn Protocol Inc) calibrated according to the manufacturer's recommendations. Standard or large cuffs were used as appropriate.

2.3. Statistical analysis

Statistical analysis was generated using SAS software, version 9.4 (SAS Institute Inc). Continuous variables were presented as mean ± standard deviation. Categorical variables were presented as number (percentage). Student t test was used to compare the value of continuous variables between study groups, and Fisher exact test was used to compare the value of categorical variables between study groups. Overall survival was assessed by Kaplan‐Meier survival analysis, with log‐rank test. Hazard ratios were assessed by the Cox proportional hazards model. This Cox model, with the Fine and Gray correction for competing risks, was used to assess hazard ratios for cardiovascular end points, with death without CHF as a competing risk. Two‐sided P values <.05 were considered statistically significant.

3. RESULTS

3.1. Study population

Overall, 410 patients were identified and included in the study: 205 in the study group who met all study inclusion criteria and a matched control group that included 205 patients. Within the study group, 139 patients (68%) had measurements of systolic BP ≥180 mm Hg and/or diastolic BP ≥110 mm Hg. Four patients from the control group were lost to follow‐up and were excluded from the analysis. Baseline characteristics of patients included in the final analysis are presented in Table 1. The average age was 67.39 ± 16.23 years in the study group and 68.47 ± 16.14 years in the control group, and 93% of the cohort was Jewish. A significant majority of the cohort had a previous diagnosis of hypertension and most patients in both the study and the control groups were taking antihypertensive medications before the ED visit. The most commonly used medications were β‐blockers (66 patients in the study group [32.2%] and 78 patients in the control group [39%]), calcium channel blockers (65 patients in the study group [31.7%] and 67 patients in the control group [33.5%]), and angiotensin‐converting enzyme inhibitors (ACEIs) (63 patients in the study group [30.7%] and 55 patients in the control group [27.5%]). Ischemic heart disease, CVA/transient ischemic attack, CHF, or dyslipidemia were similarly prevalent in the study and control groups. A preadmission diagnosis of chronic renal failure (CRF) was significantly more prevalent in the study group compared with the control group (23 patients in the study group [11.2%] and 11 patients in the control group [5.5%], P = .05), but the creatinine level at admission to the ED was not significantly different between the two groups. In the hypertensive group, cardiac symptoms, dyspnea, and weakness on admission were significantly more common than in the control group.

Table 1.

Baseline characteristics of the study cohort and clinical and laboratory characteristics associated with admission to the ED

	Elevated BP	Control	P value
No.	205	201
Age, y	67.39 ± 16.23	68.47 ± 16.14	.50
Women	106 (50.2)	105 (49.8)	.92
Prior hypertension	145 (70.7)	144 (71.6)	.91
Prior treatment for hypertension	123 (60)	132 (66)	.22
No. of drugs	1.40 ± 1.38	1.52 ± 1.36	.39
Comorbidities
Dyslipidemia	84 (41.0)	76 (37.8)	.54
Ischemic heart disease	47 (22.9)	54 (27.0)	.36
Cerebrovascular disease	27 (13.2)	23 (11.4)	.65
Congestive heart failure	6 (2.9)	13 (6.5)	.10
Chronic renal failure	23 (11.2)	11 (5.5)	.05
Admission systolic BP, mmHg	186.75 ± 23.43	135.5 ± 20.30	<.001
Admission diastolic BP, mmHg	98.26 ± 17.31	73.50 ± 13.36	<.001
Heart rate, beats per min	77.86 ± 16.46	76.96 ± 17.05	.59
Admission creatinine, mg/dL	0.97 ± 0.43	0.90 ± 0.36	.07
Symptoms
Asymptomatic	35 (17.1)	2 (1.0)	<.001
Headache	85 (41.5)	3 (1.5)	<.001
Cardiac	30 (14.6)	13 (6.5)	.01
Dyspnea	15 (7.3)	4 (2.0)	.02
Weakness	38 (18.5)	4 (2.0)	<.001
Gastrointestinal	18 (8.8)	59 (29.4)	<.001
Recurrent falls	1 (0.5)	34 (17.0)	<.001
Blurred vision	11 (5.4)	1 (0.5)	.01
Bleeding	1 (0.5)	0 (0.0)	1.00
Nonspecific symptoms	55 (26.8)	92 (45.8)	<.001
Hospitalization	42 (20.5)	73 (36.3)	<.001

Abbreviations: BP, blood pressure; ED, emergency department. Values are expressed as mean ± standard deviation or number (percentage).

3.2. Clinical outcomes in a follow‐up period of 18 months

Clinical outcomes 18 months following the ED visit are presented in Table 2. The composite primary end point (all‐cause mortality, myocardial infarction/ACS, CVA/transient ischemic attack, or hospitalization for heart failure) was similar in both the study and the control groups. A forest plot of the clinical outcomes included in the final analysis is presented in Figure 1. The only parameter that differed between study groups was hospitalization for heart failure, which was significantly more common in patients with elevated BP during the ED visit compared with patients who presented with normal BP (19 patients [9.4%] in the study group compared with 6 patients [3%] in the control group, P = .01). The diagnosis of CHF was made first by clinical parameters such as dyspnea, chest pain, respiratory deterioration, orthopnea, hypertension, desaturation, and auscultation findings suggesting pulmonary edema and based on chest x‐ray. Echocardiography was not performed in all patients. The cumulative incidence of CHF in both groups at 18‐month follow‐up is presented in Figure 2. Subgroup analysis of clinical outcomes among patients admitted to the ED with hypertensive urgency, defined as systolic BP >180 mm Hg or diastolic BP >110 mm Hg without end organ damage, compared with matched controls, at a follow‐up period of 18 months is presented in Table 3. None of the outcomes were significantly different between the two study groups.

Table 2.

Clinical outcomes during an 18‐month follow‐up period

	Elevated BP	Control	P value
Mortality	15 (7.3)	19 (9.4)	.42
ACS/MI	10 (5)	5 (2.5)	.21
CHF	19 (9.4)	6 (3.0)	.01
CVA/TIA	14 (6.9)	15 (7.5)	.80
AF	18 (8.9)	18 (9.0)	.94
Δ Creatinine	0.17 ± 0.91	0.14 ± 0.77	.73
Composite outcome	46 (22.7)	36 (17.9)	.28

Values are expressed as mean ± standard deviation or number (percentage).

Abbreviations: ACS, acute coronary syndrome; AF, atrial fibrillation; BP, blood pressure; MI, myocardial infarction; CHF, chronic heart failure; CVA, cerebrovascular accident; TIA, transient ischemic attack.

Δ Creatinine = The difference between the creatinine level after 18 months of follow‐up and the creatinine level at admission to the emergency department.

Figure 1.

Forest plot of the clinical outcomes included in the final analysis. ACS indicates acute coronary syndrome; CI, confidence interval; CHF, congestive heart failure; CVA, cerebrovascular accident; HR, hazard ratio; MI, myocardial infarction; TIA, transient ischemic attack.

Figure 2.

Cumulative incidence of congestive heart failure (CHF) in a follow‐up (FU) period of 18 months. Con indicates control group with normotension. BLP indicates study group with hypertension.

Table 3.

Subanalysis of clinical outcome among patients admitted to the ED with hypertensive urgency during an 18‐month follow‐up period

	Elevated BP	Control	P value
Mortality	10 (7.2)	13 (9.6)	.46
ACS/MI	9 (7.4)	5 (4.2)	.30
CHF	9 (7.3)	6 (5.0)	.45
CVA/TIA	9 (7.4)	11 (9.5)	.61
AF	13 (10.8)	14 (14.4)	.77
Composite outcome	29 (20.9)	28 (20.6)	.10

Values are expressed as number (percentage). Abbreviations: ACS, acute coronary syndrome; AF, atrial fibrillation; BP, blood pressure; MI, myocardial infarction; CHF, chronic heart failure; CVA, cerebrovascular accident; ED, emergency department; TIA, transient ischemic attack.

3.3. Post ED medical treatment

Within the study group admitted to the ED with elevated BP, 178 patients (88.1%) were discharged with a recommendation of continuing medical treatment for hypertension. The drugs that were in use following the ED visit are presented in Figure 3. The most commonly used drugs were calcium channel blockers (112 patients [55.5%]), compared with 138 patients (68.3%) who used ACEIs or ARBs. Overall, 166 (81%) patients in the study cohort regularly used their antihypertensive medications.

Figure 3.

Medical treatment used to treat hypertension in the community following an emergency department visit. ACEi indicates angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; BB, β‐blocker; CCB, calcium channel blocker.

4. DISCUSSION

The main finding of this study is that hospitalization for heart failure occurs more commonly during an 18‐month period in patients who present to the ED with elevated BP compared with those who present with normotension. There is limited literature on the long‐term outcome of patients admitting to the ED with elevated BP or hypertensive urgency. Our findings support the findings of Merlo and colleagues4 who evaluated the outcomes of patients visiting their general practitioner with elevated BP at 1‐year follow‐up. This study shows that those patients experienced more cardiovascular events, particularly acute CHF.

A significant majority of the cohort had a previous diagnosis of hypertension, a finding that supports evidence from previous studies assessing patients with hypertensive crises.4, 18 Perhaps hypertensive crisis occurs more frequently in patients with known hypertension and it is possible that the cardiovascular damage began long before the admission to the ED.

Although most of the hypertensive cohort was on medical treatment, we do not know whether their BP was controlled prior to the ED visit. However, because groups were matched according to hypertension diagnosis and because use of antihypertensive medications was similar between the two groups, it seems that the damage caused by preexisting hypertension was expected to be similar between the two groups.

All comorbidities did not differ between the study and control groups, serving as evidence that the incidence of hospitalization for CHF could not have been explained by preexisting illness. Particularly, a diagnosis of CHF was prevalent to a similar degree in both the study and control groups.

In contrast to the comorbidities above, the proportion of previous diagnoses of CRF was significantly higher in the study group compared with the control group. CRF is a cardiovascular risk factor and therefore may have influenced the development of CHF in the study group. Yet, since CRF is a known risk factor for mortality and ischemic heart disease, both of which occurred at a similar rate between the two groups, we believe that the disparity in prior diagnosis of CRF does not explain the higher incidence of hospitalization for CHF in the study group. In addition, the creatinine level at admission to the ED was not significantly different between the two groups, indicating that the renal status of the patients was not significantly different between the two groups at admission.

In the hypertensive group, the main findings at presentation to the ED were cardiac symptoms, dyspnea, and weakness, and these symptoms might suggest that CHF was more prevalent in patients with hypertension. However, since all patients with hypertension with target organ damage were excluded, it is less likely that there was more CHF in the hypertensive group.

In the current study, no difference in the incidence of the combined primary outcome was noted between the two groups. This finding does not support the findings of Vlcek and colleagues,15 in which the frequency of cardiovascular events was significantly higher in patients with hypertensive urgencies compared with controls. In their study, cardiovascular events were defined as ACS, CVA, atrial fibrillation, acute CHF, and aortic aneurysm in a follow‐up period of at least 2 years. In our study, we did not include atrial fibrillation in the composite outcome since this outcome is difficult to evaluate because many episodes of atrial fibrillation are asymptomatic. In addition, in their study, only patients with hypertensive urgency were evaluated, probably representing the extreme form of uncontrolled hypertension in the ED setting. Recent studies investigating the long‐term outcome of patients with hypertension also examined parameters such as aortic dissection,19 aortic aneurysm15 and peripheral arterial occlusion.4 We decided not to include these outcomes in our analysis because of their rarity. Mortality was similar between the two groups, in line with findings reported by Merlo and colleagues,4 who have shown that at 1‐year follow‐up, 5% of the cohort did not survive. One possible explanation for this finding is that patients who had elevated BP during the ED visit were followed and treated more closely. It is also possible that the short follow‐up period may have been insufficient to reveal a significant mortality difference between the two groups and longer follow‐up periods are required to investigate such a difference. In a subanalysis of clinical outcomes in patients admitted to the ED with hypertensive urgency only, none of the outcomes were significantly different between the two study groups. We believe the disparity in results is likely attributable to the smaller size of the cohort with hypertensive urgency.

Almost all patients were discharged with a recommendation for antihypertensive treatment. More than two thirds of patients were discharged on ACEIs and/or ARBs and compliance during follow‐up was high. Treatment with ACEIs/ARBs decreases the risk for CHF in patients with hypertension according to several studies,20 but because most of the cohort was treated with ACEIs/ARBs, we could not evaluate whether treatment with ACEIs/ARBs was associated with less CHF compared with treatment with other antihypertensive agents.

5. STUDY LIMITATIONS

This study has several limitations. Because of its retrospective nature, we could not follow the degree of BP control after discharge, physician visits, and other parameters that may have influenced survival. The small population size is also a limitation since the number of outcomes is scanty; however, all cases with a primary discharge diagnosis of hypertension in 2012 to 2015 were included in the analysis. Most of our study cohort (68%) had BP ≥180/110 mm Hg, but patients with elevated BP not defined as having a hypertensive urgency were purposely included in the analysis as the outcome of such patients was not previously reported. If patients with only hypertensive urgencies were included the analysis, we may have seen a difference in other outcomes between the two groups.

We collected data regarding previous comorbidities of our study population, including heart failure. However, data on left ventricular function and B‐type natriuretic peptide levels from the ED visit were not available since these evaluations are not routinely performed in patients admitted to the ED with hypertension. Therefore, it is possible that some cases of asymptomatic heart failure may have been missed on the initial analysis. Yet, because we matched the groups for a history of hypertension and because comorbidities were similar between the two groups (except for CRF), we have no reason to believe that the prevalence of HF differed between the study groups.

6. CONCLUSIONS

Elevated BP during an ED visit is associated with increased risk for hospitalization for CHF during an 18‐month follow‐up period. These patients require strict follow‐up and evaluation for complications. Whether treatment with ACEIs/ARBs is more appropriate for these patients to prevent this outcome requires further study.

CONFLICT OF INTEREST

The authors have no conflicts of interest to declare.

Ayalon‐Dangur I, Rudman Y, Shochat T, Ofer‐Shiber S, Grossman A. Elevated blood pressure during emergency departments visit is associated with increased rate of hospitalization for heart failure: A retrospective cohort study. J Clin Hypertens. 2018;20:98–103. 10.1111/jch.13155