Sunday, January 13, 2013

Severely Elevated Blood Pressure in the ED

The Gist:  Hypertension is common within the ED population but the workup and treatment of the asymptomatic patient with a scary high blood pressure (BP) is variable.  Evaluate each patient independently but patients with asymptomatic, severely elevated BPs do not routinely need investigations (lab, ECG, UA) or urgent medication. If an anti-hypertensive is used in this population, use oral medications and avoid hydralazine, clonidine, and nitrates.  Arrange good follow up with a PCP.  Listen to these: Dave KarrasRob Rogers, EMBasic, and Chad Kessler.

The FOAM world likely practices similarly to the experts listed in the above talks, but depending on the ED and attending I'm working with, high BPs in the ED are managed in very different ways.  At one institution, labetalol, hydralazine, and clonidine were liberally administered despite symptomatology.  In another, we discharged patients without workup or treatment with pressures >190/100.   I've learned that EM practice varies widely, but what's the rationale and evidence behind these disparate approaches?

Whenever I have clinical questions, I do some basic review in a core text (i.e. Tintinalli or Rosen).  However, I often first check FOAM sources so that I can get information quickly that may be a little more updated. 

What are we worried about? (not the focus of this post)
Hypertensive emergency - Rapid, progressive decompensation of vital organ, characterized by  evidence of acute dysfunction in cardiovascular, neurologic, or renal systems, caused by an inappropriately increased BP. There is no numerical definition (1).  These are the sick patients who need aggressive treatment.  These are:
  • Acute pulmonary edema
  • Aortic dissection
  • Pre-eclampsia
  • Subarachnoid hemorrhage
  • Hypertensive encephalopathy
  • Acute kidney injury
  • Intravenous blood pressure reduction may be indicated in other medical scenarios such as acute coronary syndrome, acute ischemic stroke, and intracerebral hemorrhage.  
  • Some medical conditions can cause a severely elevated blood pressure, so keep these in mind when evaluating a patient:  thyroid storm, cocaine, autonomic dysreflexia, etc but usually these are accompanied by symptoms and sickness.
What do most of the patients actually have?
Severe Asymptomatic Hypertension -  previously known as "Hypertensive Urgency" (SBP > 180 mmHg or DBP > 110 mmHg), is more appropriately an outpatient term - they need their blood pressure controlled urgently (borrowed from this great Emergency Board Review podcast).
  • Hypertension diagnosis requires repeated measurements over time, so a better term in the ED may be severely elevated blood pressure (from EMA 08/08).  

  • The ED is not the ideal place to gauge an individual's baseline blood pressure
    • Can be elevated due to pain, stress, anxiety.  This retrospective, observational study by Svenson demonstrates that 40% of the pediatric patients had blood pressure readings that qualified as 40% of kids have hypertension?  Maybe soon in the United States.  In fact, about half of patients may not have hypertension in the outpatient setting (Tanable et al). 
  • The patient may unknowingly live with this blood pressure.  As mentioned in the Tanable study above, the other half of patients in the ED with elevated blood pressures may have persistently elevated blood pressures in the outpatient setting.  This point seems to be born out across many studies, where the numbers range from 25-50% depending on the rigor of follow up and methodology (2).  
But don't they have symptoms if they're in the ED?
  • Some providers refer patients to the ED for very high BPs or the patient may be scared of the number at home BP checks and desire further evaluation.  
  • What about headaches, dizziness and epistaxis?  Ok, the epistaxis caused by hypertension myth has been debunked and most people know this.  
    • Which came first, the symptom or the elevated blood pressure?  Pain, discomfort, or stress may engender an elevated blood pressure.  On other occasions, serious pathology may result from high blood pressure, causing symptoms.
    • Height of blood pressure does not correlate with symptoms often ascribed to blood pressure, such as dizziness, headache, etc (Karras 2005).   
    • Headaches in the ED patient with severely elevated blood pressure seem to cause the most concern amongst practitioners.  Emergency Physicians must detect the badness - things like subarachnoid hemorrhage and hypertensive encephalopathy.  However, headache is a common patient complaint in the ED, accounting for 1-4% of ED visits (3), elevated blood pressure is also common in the ED, and the majority do not have intracranial pathology (23).  This is gray, non-evidence based area but history and physical probably play a big role in differentiating whether a patient's headache is concerning for hypertensive "end-organ" pathology, which would warrant imaging.  Is the patient on anti-coagulants?  Do they have a neuro deficit or vision changes?  Altered mental status?  Do they look sick?  What does dizzy really mean?  Perhaps one could even check for increased ICP with a quick bedside ultrasound.   
      • Note: The 3.8% incidence of intracranial pathology in headache touted in Tintinalli and many papers is from a study in which 459 records from 1720 patients with headache were randomly selected and reviewed (Ramirez).  The more recent study is a retrospective records review, much larger in caliber and showed an incidence of ~2% (of 2.1 million) for any pathologic process, but is limited by the follow up and that it's a database review, relying on coding (Goldstein).  
Workup - History and physical may yield an easy culprit.  For example, patient's often fail to take their medications, as evidenced by 42% of the hypertensive cohort in this study.

No need for testing in truly asymptomatic patients (Kessler talk).  The newly updated ACEP clinical policy supports this recommendation (Level C). This means, no creatinine or screening for "end organ damage" unless a patient has poor follow up and you think that the result of the test will affect disposition (ie admission).  
  • Karras et al study demonstrated that in an asymptomatic hypertensive ED cohort subjected to a battery of tests, 52% had abnormal results.  Management was changed in only 6% (n=7) of patients (could have been anything from more testing to ICU admission). 
    • Limitation:  Small study that excluded patients with acute illness and "emergent" complaints/triage categories.
If you know you're going to start an anti-hypertensive, check a BMP (Kessler talk).  This is less evidence based but makes some sense.
  • Look for creatinine as a surrogate for renal function (relevant to diuretics, ACEIs) and potassium (before starting an ACEI).
    • Tintinalli suggests that a urinalysis, looking for hematuria and proteinuria, may be used instead of a BMP to check for renal insufficiency (Ch. 61).
What about searching for markers of organ damage?
  • May indicate the complications of chronically elevated BP such as elevated creatinine, proteinuria, or hypertensive heart disease, which may be prevalent in the asymptotic population (Levy study).  As chronic problems, these necessitate chronic treatment. 
  • CXR and ECG have poor sensitivity and are unlikely to change management (Bartha.  Limitation: study 35 years old, outpatient setting, n=116, changed management in 2/116 cases).
Don't routinely use anti-hypertensives in the asymptomatic ED patient.  This is the meat of the "controversy" in hypertension in the ED.  The evidence base and consensus agree that asymptomatic patients with severely elevated blood pressures don't warrant acute reduction.  The ACEP guidelines (Level C) do not recommend treating patients who are lack evidence of end-organ damage.
  • This can decrease cerebral perfusion pressure in patients who have a chronically elevated blood pressure, which can lead to a stroke/cerebral hypoperfusion.  This is bad. Mechanism:  The cerebral vasculature has amazing autoregulation, maintaining the brain's blood flow despite changes in mean arterial pressure (MAP).  In chronic hypertension, this mechanism is shifted so that auto-regulation occurs at higher MAPs (lost at a higher MAP as well) leaving them unable to tolerate a rapid return to normal blood pressure.
  • Lowering the blood pressure in the ED doesn't necessarily translate into patient benefit.  It may make us feel better, but does it really help the patient?  
    • Gallagher editorial points out that the only real evidence that addresses this question stems from a 1967 VA study. 
So, why does this practice remain in some areas?
  • Many patients don't have any other point of contact with the healthcare system.  
    • Limitation:  This doesn't necessarily mean a patient can fill a prescription or will establish longitudinal care to continue receiving scripts and medical oversight.
  • Habit.  We practice how we're taught and within the local culture.  Many EM texts still recommend treatment of asymptomatic high BPs in the ED.  
  • Patient expectation:  One study demonstrates that the likelihood that a patient is more likely to receive an anti-hypertensive if their chief complaint is "high blood pressure" with an OR 5.6 (95% CI 2.0 -15.3).  When busy, it's tempting to easily satiate the patient with a temporizing measure.  
Prescribing an anti-hypertensive in the ED:
  • Use oral anti-hypertensives (Kessler talk)
    • If patient receives IV anti-hypertensives, they should probably be admitted (Rob Rogers)
  • Don't give clonidine, hydralazine, nitratesThese are unpredictable and if you're giving these, your patient is probably sick (and therefore, not suffering from asymptomatic severe hypertension) (Rogers, Kessler)
  • Drop pressure gradually without an aim to achieve a normal BP (ACEP guidelines)
  • The JNC VIII guidelines recommend: Hydrochlorothiazide, ACEI, ARB, or calcium channel blocker for non-black patients and hydrochlorothiazide or calcium channel blocker for black patients is still first-line per JNC VII for most patients.  
    • See this ERCast episode for more anti-hypertensive prescribing pearls
  • In patients that have poor follow up, ACEP recommends considering initiating outpatient oral anti-hypertensive therapy (Level C).
Refer the patient for outpatient follow up.  The ACEP guidelines give this a level C recommendation for those with persistently elevated BP readings.  
  • A good idea as literature base show that many (~50% in some studies) will actually have hypertension (Umscheid et al).  
  • The patient may very well develop (or already have) the sequelae of chronic hypertension, which can cause significant morbidity. 
The blood pressure may fall on its own.  
  • This study by Cienki et al looked at a small sample of ED patients with high blood pressure and took measurements at 10 minute intervals to examine the fall of these values.  They found that the blood pressure did typically fall significantly in the cohort with BP >160/100 (the population we're really talking about here).  
    • Limitations:  The study was really small and did not include patients who ingested caffeine/stimulants so while the conclusion is reasonable, it's not entirely applicable to the typical ED population.  
Other good resources:
Hypertension and the Emergency Physician - post by Dr. Reuben Strayer
EB Medicine June 2010 review
Commentary/Review by Drs. Shayne and Pitts from Emory

Updated 01/21/2014


  1. Learned a lot from this article, thanks for posting it Lauren!

  2. This comment has been removed by the author.