Retrospective review of the use of as-needed hydralazine and labetalol for the treatment of acute hypertension in hospitalized medicine patients

Abstract

Background:

The aim of this study was to evaluate the use of as-needed (PRN) labetalol and hydralazine [intravenous (IV) or oral] in hospitalized medicine patients for the treatment of severe asymptomatic hypertension and to examine the potential negative outcomes associated with their use.

Methods:

The electronic health record of 250 medicine patients hospitalized at the University of Colorado Hospital between November 2014 and April 2016 who received at least one dose of PRN IV or oral hydralazine or labetalol were retrospectively reviewed. The primary outcome was to describe the use of PRN antihypertensive medications in this population.

Results:

A total of 573 PRN doses of antihypertensive medication were administered. Oral hydralazine was the most common (521 doses, 90.9%). A total of 36% of PRN administrations were given for a systolic blood pressure (SBP) <180 mmHg and diastolic blood pressure (DBP) <110 mmHg (cut-point for acute severe hypertension). No serious adverse events were related to PRN antihypertensive administration. Despite receiving at least one PRN antihypertensive medication during hospitalization, 40.8% of patients were not continued on their home antihypertensive medication(s) while hospitalized, and 62.4% of patients did not have their home regimens intensified at discharge.

Conclusion:

As-needed oral hydralazine is frequently prescribed for acute blood pressure lowering with administration thresholds often less than what are used to define acute severe hypertension. Many patients are prescribed PRN antihypertensive medication instead of being continued on their home regimens, and most patients do not have the intensity of their home regimens increased. Providers need to be educated about the use of PRN antihypertensive medication for the management of severe asymptomatic hypertension in the hospital setting.

Introduction

About one-third of adults in the United States have chronic hypertension, defined as a systolic blood pressure (SBP) ⩾140 mmHg, or a diastolic blood pressure (DBP) ⩾90 mmHg, or both, on repeated examination.¹ Providers treat chronic hypertension based on comprehensive, evidence-based guidelines.^1–3 Acute severe hypertension (hypertensive crisis) is frequently defined as a SBP ⩾180 mmHg or a DBP ⩾110–120 mmHg diastolic.^2–5 This is further subdivided into hypertensive emergency, in which signs and symptoms of end-organ damage are present or hypertensive urgency, also called severe asymptomatic hypertension, in which there is no evidence of end-organ damage.^4,6 The management of acute severe hypertension is not addressed in newer guidelines and is only briefly mentioned in older ones.^1–3,7 Consequences of chronic uncontrolled blood pressure are well known; however, data on the consequences of severe acute blood pressure elevations, especially in the asymptomatic patient, are lacking. In addition, no studies have demonstrated a benefit of acute blood pressure lowering, even in patients with hypertensive emergency, but the potential exists for adverse consequences if blood pressure is decreased too rapidly.⁸

The prevalence of hypertension in the hospital setting is as high as 50%.⁹ There are several reasons a patient may have acute elevations in blood pressure while hospitalized such as: uncontrolled pain from trauma/surgery or a medical condition, volume overload, anxiety, new medications that increase blood pressure, or having home antihypertensive medications held.^10,11 Since there is a lack of data guiding the management of severe elevations in blood pressure, these patients are frequently and inconsistently managed with as-needed (PRN) antihypertensive medications.

Several studies have demonstrated the inappropriate use of PRN intravenous (IV) hydralazine and labetalol for the management of severe blood pressure elevations in hospitalized patients.^10–13 In these studies, IV antihypertensives were frequently administered to patients that did not meet criteria for acute severe hypertension. In addition, patients were often not continued on their home antihypertensive medications prior to administration of a PRN IV antihypertensive agent.

The present study evaluated the use of PRN hydralazine and labetalol both IV and oral in hospitalized internal medicine patients with severe asymptomatic hypertension. It was hypothesized that medications used to treat severe asymptomatic hypertension, in the inpatient setting, are inconsistently given based on drug, dose, frequency, and blood pressure readings before administration. The purpose of this study was to describe the use of these agents, assess potential harm associated with acute treatment of severe asymptomatic hypertension, and describe changes to the patients’ antihypertensive medication regimens during hospitalization and on discharge.

Methods

This single-center, retrospective, medical chart review was conducted at the University of Colorado Hospital (UCH), CO, USA, a 648-bed urban academic hospital that utilizes 100% computer prescriber order entry and bar code technology to document medication administration. Study participants were identified from generated reports of internal medicine patients who had been charged for at least one dose of PRN IV or oral hydralazine or labetalol. These reports were generated using Clarity, an analytical database that is extracted from information that resides in EPIC (Madison, WI, USA), UCH’s electronic medical record. A convenience sample of 250 patients was prospectively chosen. Patient electronic medical records (EMRs) were retrospectively reviewed starting on 22 April 2016 until 250 patients were included. Patients included in this study received PRN antihypertensive medication administration between 15 November 2014 and 22 April 2016. The Colorado Multiple Institutional Review Board approved this study (COMIRB # 14-2110). Informed consent was waived.

Patients

Patients were included in this study if they were 18–89 years of age, admitted to a UCH internal medicine service, and had at least one documented administered dose of PRN IV or oral hydralazine or labetalol. Patients were excluded if they were admitted to the intensive care unit, pregnant, had a cerebrovascular accident within the last 6 months, admitted for hypertensive emergency or for conditions associated with hypertensive emergency such as cerebral hemorrhage, hypertensive encephalopathy, and aortic dissection or aneurysm.

Data collection and study outcomes

The following data were collected from the EMR for all patients: basic demographic information including age, sex, race, body mass index (BMI), renal function, comorbidities, antihypertensive regimen (if any) prior to admission, changes to oral antihypertensive therapy during admission, and antihypertensive medications ordered on discharge. We also collected detailed information on the PRN antihypertensive medication order such as: drug (hydralazine or labetalol), dose, route, frequency, blood pressure threshold, and day and time of administration. To analyze the outcomes associated with PRN antihypertensive therapy, the following data were collected: pretreatment blood pressure values (blood pressure immediately prior to antihypertensive medication administration), changes in blood pressure within 6 h after documented PRN administration, evidence of a new stroke, acute kidney injury (AKI) or acute coronary syndrome (ACS) within 24 h, and report of a patient experiencing a fall within 12 h of administration of a PRN antihypertensive medication. In cases where post 6 h blood pressure data were not available, the data were not included in the analysis. Other information collected included documentation of nursing or physician assessment of the presence of end-organ damage.

We defined an adverse outcome as a 25% decrease in SBP or DBP within 6 h of PRN antihypertensive medication administration, a new stroke, AKI, or ACS within 24 h of administration, or a report of a patient experiencing a fall within 12 h of administration. The chosen decrease in blood pressure is consistent with recommendations on blood pressure lowering in the setting of hypertensive emergency.³ This definition is also supported by studies of the possible negative effect of lowering blood pressure too rapidly and reflects criteria used in other trials.^12,13

Statistical analysis

Descriptive statistics were performed for all variables. Continuous data were summarized using means and standard deviations. Categorical variables were summarized as counts and percentages. Binary logistic regression was used to model the likelihood of having home antihypertensive medications intensified at discharge. A patient’s home antihypertensive medication regimen was considered intensified if their antihypertensive medication dose(s) was increased, another blood pressure medication was added, or treatment-naïve patients were started on a medication. For the logistic regression, gender was a binary variable of female or male. Age groups were categorized into three values: 18–45, 46–65, and 66–89 years. Guidelines from the World Health Organization were used to categorize BMI as: underweight, normal weight, overweight, or obese. Obesity classes 1, 2, and 3 were confined to a single category (obese) for this study. Race and ethnicity were combined into a single classification with categories of: non-Hispanic White, Hispanic White, Black or African-American, or other. Statistical analysis was performed using SAS software, version 9.4. (SAS Institute Inc., Cary, NC, USA).

Results

Baseline characteristics

During the study period, 381 internal medicine patients were charged for at least one PRN dose of IV or oral hydralazine or labetalol. After applying the inclusion/exclusion criteria, 250 patients were included in this study (Figure 1). Characteristics of the patients are shown in Table 1. Among our study cohort, patients had an average age of 55.6 ± 15.5 years, and approximately half (49.2%) were female. Most patients had a history of hypertension. The majority of patients (82.4%) were prescribed various antihypertensive medications prior to admission; the most common were renin-angiotensin blockers (46.8%) and beta-blockers (46.0%).

Figure 1.

Patient allocation and changes to antihypertensive regimens at discharge.

BP, blood pressure; CVA, cerebrovascular accident; IV, intravenously; PO, orally; PRN, as-needed.

Table 1.

Baseline characteristics of the study population.

Demographics and characteristics	n = 250
Mean ± SD age, year	55.6 ± 15.5
Females, no. (%)	123 (49.2)
Race, no. (%)
White	147 (58.8)
Black	84 (33.6)
Past medical history, no. (%)
Hypertension	216 (86.4)
Diabetes	128 (51.2)
Chronic kidney disease	75 (30.0)
Coronary artery disease	46 (18.4)
Obstructive sleep apnea	28 (11.2)
Stroke/TIA	22 (8.8)
Systolic heart failure	23 (9.2)
Diastolic heart failure	20 (8.0)
Renal transplant	26 (10.4)
ESRD on dialysis	33 (13.2)
Anxiety	35 (14.0)
Post-MI	26 (10.4)
Pulmonary hypertension	10 (4.0)
Atrial fibrillation	20 (8.0)
Medications prior to admission, no. (%)
Renin-angiotensin blockers	117 (46.8)
Beta-blockers	115 (46.0)
Calcium antagonists	74 (29.6)
Loop diuretics	48 (19.2)
Thiazide diuretics	27 (10.8)
Hydralazine	20 (8.0)
Aldosterone antagonists	17 (6.8)
Other	25 (10.0)
No BP medications prior to admit	44 (17.6)
Mean ± SD BP medications prior to admission, no.	1.8 ± 1.3

BP, blood pressure; ESRD, end-stage renal disease; MI, myocardial infarction; SD, standard deviation; TIA, transient ischemic attack.

As-needed antihypertensive medication administration

A total of 573 doses of PRN IV or oral hydralazine or labetalol were administered to the 250 patients. Oral hydralazine was the most commonly administered antihypertensive agent at 90.9% of doses compared with IV hydralazine at 5.4%, oral labetalol at 3.0%, and IV labetalol at only 0.7% of doses. A total of 74% of the administered oral hydralazine doses were 25 mg and 50 mg. The most common dosing frequency for the PRN antihypertensive medication orders was Q6 hours (44.3%) and Q8 hours (48.3%). Almost half (46.4%) of the patients received one dose of PRN antihypertensive medication, 59 (23.6%) received 2 doses, 31 (12.4%) received three doses, and 44 (17.6%) received >3 doses. There was one patient that received 12 doses. The percent of PRN antihypertensive doses administered during the afternoon (39%) and evening shifts (37%) were higher than the morning shift (24%).

Blood pressure threshold parameters were included in all of the PRN antihypertensive orders. A total of 91% of the PRN doses given met the blood pressure order threshold prior to administration. Table 2 summarizes the blood pressure readings prior to PRN antihypertensive administration. Overall, 36% of PRN administrations were given when the SBP was <180 mmHg and DBP was <110 mmHg, which is below the blood pressure cutoff for acute severe hypertension. The mean blood pressure that triggered administration of a PRN antihypertensive medication did not differ among shifts (morning shift = 182/93 mmHg, afternoon shift = 184/94 mmHg, and evening shift = 182/93 mmHg).

Table 2.

Blood pressure readings prior to PRN antihypertensive administration.

BP readings (mmHg) prior to administration	Number of administrations out of 573
SBP <150	4
SBP 150–159	17
SBP 160–169	56
SBP 170–179	127
SBP 180–189	156
SBP 190–199	116
SBP ⩾200	60
SBP <180, but DBP ⩾110	37

The bolded line in the middle indicates acute severe hypertension cutoff.

BP, blood pressure; DBP, diastolic blood pressure; PRN, as-needed; SBP, systolic blood pressure.

Blood pressure outcomes

A total of 6 antihypertensive administrations did not have a 6-h post blood pressure, so all results are out of 567 administrations. The mean pretreatment blood pressure was 183/93 ± 14/13 mmHg. During the 6 h following administration of PRN antihypertensive medication, the blood pressure was reduced by a mean of 29/12 ± 21/12 mmHg.

Figure 2 describes the blood pressure changes within 6 h of PRN antihypertensive administration. SBP or DBP decreased by <10% in 13% of administrations and by 10–25% in 50% of administrations. A total of 22% of administrations were associated with overtreatment defined as a greater than 25% reduction in blood pressure. There were 3% of administrations (18/567) that resulted in an increase in blood pressure. The average SBP increase was 7.2 ± 5.1 mmHg (range 1–19 mmHg). The average increase in DBP was 4.5 ± 3.5 mmHg (range 1–16 mmHg). Since all 18 administrations were in 18 different patients, 18/250 or 7.2% of patients had an increase in blood pressure after receiving an antihypertensive agent.

Figure 2.

Blood pressure changes within 6 h of as-needed antihypertensive medication administration (n = 567 administrations).

Adverse event outcomes

There were no serious adverse events related to PRN antihypertensive administration including no documented cases of a new stroke, AKI, or ACS. A greater than 25% reduction in blood pressure within 6 h of PRN antihypertensive medication administration was observed in 122 out of 567 administrations (21.5%). There was one patient who received oral hydralazine that experienced a headache and one patient who received oral hydralazine that experienced dizziness post-administration. Also, one patient fell within 2.5 h of oral hydralazine administration. This patient had a history of a previous fall and was on two other medications associated with falls. An assessment of end-organ damage was not documented in any patient prior to administration of a PRN antihypertensive medication.

Changes to home antihypertensive therapy

Patients were on an average of 1.8 ± 1.3 antihypertensive medications prior to admission and on an average of 2.0 ± 1.2 antihypertensive medications at discharge. Figure 3 summarizes the data on continuation of home medications during hospitalization. A total of 84 (40.8%) of the 206 patients that were prescribed chronic antihypertensive medications prior to admission were not continued on these home medications while hospitalized. Of the 122 patients that were continued on their home antihypertensive medications while hospitalized, 39 (31.9%) were started on these medications 2 days or more after admission. On discharge, 94 patients (37.6%) were newly started on an antihypertensive medication or their home antihypertensive medications were intensified (Figure 1).

Figure 3.

Continuation of home medication (n = 206 patients).

The number of PRN doses (p = 0.013), home medication continuation during hospitalization (p < 0.001), and previous diagnosis of hypertension (p = 0.023), after adjusting for age, race/ethnicity, BMI, prior medications, and previous diagnosis of diabetes, stroke, chronic kidney disease, and coronary artery disease, were all found to be significant predictors of having prescribed home antihypertensive medications intensified at discharge. For every one additional PRN antihypertensive medication administration, patients were 25% more likely to have their home medications intensified at discharge [odds ratio (OR), 95% confidence interval (CI): 1.25, 1.05–1.50]. Patients who did not continue their home antihypertensive medications during their hospital stay were 70% less likely to have their home antihypertensive medications intensified at discharge (0.30, 0.14–0.61). Additionally, patients without a previous medical history of hypertension were 77% less likely to have their antihypertensive medications intensified at discharge (0.23, 0.07–0.80).

Discussion

To date, no study has examined the use of PRN oral antihypertensive medication for the management of acute severe asymptomatic hypertension in hospitalized internal medicine patients. Our data indicate that PRN oral hydralazine is commonly prescribed to manage acute severe asymptomatic blood pressure elevations in this setting. Oral hydralazine is not recommended for the treatment of severe asymptomatic hypertension by several resources due to its unpredictable pharmacokinetics and blood pressure lowering, as well as the risk for reflex tachycardia.^14,15 Our data support this statement as evidenced by the interpatient variability in blood pressure changes within 6 h of PRN antihypertensive administration. While selected short-acting oral antihypertensive medications (e.g. labetalol, clonidine and captopril) are recommended to achieve blood pressure reductions in patient with hypertensive urgency, it is important to emphasize that these recommendations do not endorse the administration of the medications as-needed.^3,16 Instead, single-doses of these medications, along with adjustment of the patient’s chronic antihypertensive therapy, are recommended to achieve adequate blood pressure reduction over hours to days depending on the patient’s risk for end-organ damage.¹⁷

This study confirms our hypothesis that PRN antihypertensive medication used to treat severe asymptomatic hypertension in hospitalized internal medicine patients is inconsistently prescribed. While oral hydralazine was commonly prescribed, the dose, frequency, and blood pressure thresholds for administration varied widely. In addition, assessment of symptoms or end-organ damage was not documented prior to the administration of any of the PRN antihypertensive doses. Our study is consistent with other studies that have examined the use of PRN IV antihypertensive medications. These studies have also documented a wide variability in dosing and the lack of proper patient evaluation prior to antihypertensive medication administration.^5,10–13 This variability in administration practices may be attributed to the paucity of literature guiding healthcare providers on the management of severe asymptomatic hypertension in the hospitalized patient. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure states that the term ‘hypertensive urgency’ has led to overly aggressive treatment of patients with severe, uncomplicated hypertension.³ Newer guidelines do not address the treatment of hypertensive emergencies and urgencies, and other, older guidelines have vague recommendations for the management of hypertensive urgencies.^1–3,7 It is important to treat patients based on a thorough assessment of the cause of their hypertension and avoid aggressively dosing with IV or oral antihypertensive agents to rapidly lower blood pressure to simply treat the numbers.

No serious adverse events occurred in this study. Other studies have reported adverse events due to overzealous blood pressure treatment including hypotension, dizziness/lightheadedness, need for IV fluid replacement, and the need to hold oral blood pressure medications.^12,13 This could be due to the fact that the large majority of our patients received oral antihypertensive medication compared with the use of IV antihypertensive medication in the other studies.

It seems that healthcare providers prescribe PRN antihypertensive medication based on their beliefs that acutely lowering elevated blood pressure is important. A survey assessing the attitudes and practices of resident physicians regarding hypertension in the inpatient setting found that 44% of respondents would treat acute asymptomatic, moderately elevated blood pressure with either an IV or oral antihypertensive agent.¹⁸ Intravenous or oral hydralazine was preferred by 50.8% of respondents. It is important for health care providers to be educated on the lack of benefit, and possible increased risks of acute blood pressure lowering in the hospitalized patient and the downsides to PRN administration of these medications. There are no studies that demonstrate a benefit from acute blood pressure lowering in a patient with hypertensive urgency or severe uncontrolled hypertension.⁸ There are; however, data suggesting potential risks associated with aggressive blood pressure lowering.^4,10,19 Rapid blood pressure lowering may lead to inadequate tissue perfusion and the risk for kidney, cerebral, or myocardial ischemia.^4,17,19 PRN medication administration, based only on a blood pressure cutoff, is not adequate as it may not allow for appropriate assessment of the presence of or risk of end-organ damage and does not address the blood pressure issue long-term.¹⁷

The prevalence of uncontrolled hypertension in the inpatient setting, defined as a blood pressure ⩾140/90 mmHg, ranges from 23.8% to 72%.²⁰ The prevalence of severe asymptomatic hypertension is unknown. There are several possible reasons for severe asymptomatic elevations in blood pressure in the hospital setting. In our study, the majority of patients had a previous diagnosis of hypertension and many were not continued on their home antihypertensive medications while hospitalized. Severe blood pressure elevations in these patients may have been due to uncontrolled chronic hypertension or having home antihypertensive medications held. Patients without a previous diagnosis of hypertension may have had undiagnosed chronic hypertension that was now being detected with routine blood pressure monitoring. Other possible causes for acute severe blood pressure elevations in hospitalized patients include uncontrolled pain from a medical condition, trauma or surgery, acute volume overload, anxiety, or newly started medications known to increase blood pressure.^10,11,17 We were unable to adequately assess these possible causes in our study.

Despite receiving doses of PRN oral antihypertensive medication for elevations in blood pressure, the majority of patients had no changes to their home antihypertensive medications, and in some cases decreased intensity of their home antihypertensive medications. Our study did find several predictors of having patient’s home antihypertensive medications intensified at discharge. The more doses of PRN antihypertensive medication a patient received, the greater the likelihood of them having their antihypertensive medications intensified on discharge. Many patients have uncontrolled hypertension in the outpatient setting.²¹ It is possible that many of these patients who had antihypertensive regimens started or intensified at discharge had uncontrolled blood pressure prior to admission which warranted the need to intensify their regimens at discharge. Severely elevated blood pressure while hospitalized, should be a red flag to healthcare providers that a patient’s long-term blood pressure may be poorly controlled.

Patients were less likely to have antihypertensive medications intensified at discharge if they were not continued on their home antihypertensive regimen while inpatient, or if they had no previous medical history of hypertension. Home antihypertensive medications are frequently held due to the belief that short-acting PRN blood pressure medications are safer than longer acting blood pressure medications in hospitalized patients. This practice can lead to acute severe elevations in blood pressure due to a lack of chronic blood pressure management. These acute blood pressure elevations are especially true if clonidine or beta-blockers are stopped abruptly.²² The best management is to reinstitute the patient’s home antihypertensive medications and have them follow up with their primary care provider to see if intensification is needed. In a patient with no previous history of hypertension, it is reasonable to send a patient home with no antihypertensive medication and have them follow up with their primary care provider since their blood pressure may have been elevated due to other factors (i.e. pain, anxiety, volume overload).

This study reveals that treating severe asymptomatic hypertension in the inpatient setting is common. This practice is not based on guidelines or evidence of improved outcomes. Inpatient providers need to be educated on the proper management strategies for patients with severe asymptomatic blood pressure elevations.¹⁷ When a patient’s blood pressure is elevated to a SBP ⩾180 mmHg or DBP ⩾110, the patient should be assessed for the presence of, or risk for, end-organ damage. The patient should be evaluated for any underlying causes of elevated blood pressure (i.e. inadequately controlled pain or anxiety, abrupt stop in antihypertensive medication, fluid overload, etc.) and treated appropriately. If a patient is at risk for end-organ damage, it is reasonable to give a one-time short-acting antihypertensive medication such as labetalol, captopril, or clonidine. If a patient is not at risk for end-organ damage and all underlying causes have been ruled out, the patient should be observed closely and intensification of the patient’s existing home antihypertensive regimen, or the start of a long-acting antihypertensive regimen for treatment-naïve patients, should be considered. It is important to ensure that patients who are not controlled while hospitalized receive adequate follow up with a primary care provider after discharge. Multidisciplinary teams can work together to ensure patients have proper blood pressure treatment and follow up throughout a patient’s hospital stay and at discharge.

This study had multiple limitations. Since this is a single-center study, it is unknown if these results are consistent with other urban academic medical centers. However, our results are consistent with other small, single-center studies from different geographical areas. While we only evaluated patients admitted to our internal medicine services, we felt a study focusing on this population was needed. Previous studies either looked at all non-intensive care unit patients or focused specifically on surgical patients. Since this was a retrospective study, the accuracy and completeness of extracted data cannot be verified. In addition, we were unable to retrospectively address the clinical decision making behind the choice of hydralazine or labetalol for the treatment of acute hypertension and why the healthcare providers did not continue home medications. We did not confirm if medication reconciliation had occurred for those patients in which home medications were not continued. We chose to only evaluate the PRN use of hydralazine and labetalol due to our experience that they are more commonly prescribed PRN at our institution. Other short-acting antihypertensives were not evaluated, which can be considered a limitation of this study. Outcomes measured were all short-term and did not include important clinical outcomes such as long-term blood pressure management, cardiovascular events, readmissions rates and mortality.

Conclusion

Even though oral hydralazine is not recommended, it is commonly used for acute blood pressure lowering in hospitalized internal medicine patients. This study revealed that PRN blood pressure orders of hydralazine and labetalol are inconsistently prescribed, and blood pressure thresholds for administration of PRN antihypertensive medication are often lower than what is used to define acute severe hypertension. Many patients are prescribed PRN antihypertensive medications instead of being continued on their home antihypertensive regimens while hospitalized. Despite receiving PRN antihypertensive medication while an inpatient, the majority of patients did not have the intensity of their home antihypertensive medications increased. Inpatient providers need to be educated about the use of PRN blood pressure medications and appropriate strategies for the management of severe asymptomatic hypertension in hospitalized medicine patients.