sinaiem.org

This month in Journal Club, we continued our theme of prognostication papers as Corey reviewed two recent ones from Latha G. Stead’s group at Mayo.  One paper ran last year and is called “Initial Emergency Department Blood Pressure as Predictor of Survival After Acute Ischemic Stroke” (Neurology 2005, 65:1179:1183). The second paper is called “Impact of Acute Blood Pressure Variability on Ischemic Stroke Outcome” (Neurology 2006:66:1878-1881).

This is something I frankly hadn’t spent much time thinking about — all the emphasis in stroke guidelines and tPA admin has been about getting BP down into a safe range, not worrying about whose BP is too low. But the big result from the first paper was that a diastolic of < 70 mmHg, a systolic less than 155, or an MAP of less than 100 mmHg was associated with higher mortality at 90 days than those with higher BPs (even after adjusting for age, gender, NIHSS score, etc). The worst relative risk (RR) was for a diastolic less than 70; RR = 2.2 in that case, which the authors find is actually worse than the RR of having a diastolic over 105 (RR=1.9… How about that).   

 The second paper reported that patients who died within 90 days of ischemic stroke had greater swings in diastolic BP within their first 180 minutes in the ED compared to those who lived (about 44.5 mmHg max diastolic change vs 25 mmHg, or 47 mmHg max systolic change vs 30 mmHg in those who lived). Direction (rise or fall in BP) didn’t matter. This variability was an independent prognostic factor of age, stroke severity, and gender. Also interesting, though of course the authors make no effort to account for interventions (were they getting too much antihypertensives? Heck, was tPA given or withheld?) Most authors print comparison tables to show that their study arms are sufficiently similar; these authors do not.

Both papers derive their data from the same nonconvenience sample of midwesterners, though why it took them two and a half years to accumulate 681 stroke patients — when they claim they see 500/yr — is beyond me. They whittled their numbers further for the purposes of followup and subgroup analysis (the second paper is a subgroup analysis of this dataset) but their followup length and electronic database tracking was very thorough. Death is slightly more objective an unbiased than NIHSS scores, but not to those experienced with using the scale.

The authors put the patient’s risk of death within 90 days into a “null Cox model” to “calculcate the martingale residual for each patient.” I don’t know what this means but helped them generate a scatter plot that frankly looks wrong, but in any case was way over our heads. Even our programs MPH’s were unfamiliar with these manipulations. But they fitted these very skewed scatterplots with a spline curve and found blood pressure values for which mortality was high, generating the results summarized above.

As for applying these results to our setting, well, this papers make no claims to impact management. What we can do is follow NINDS, as always, watch for new guidelines from AHA and new data from this study group. Until then, patients with 140/60 ought to be considered sicker to us, worthy of a fluid bolus and alerting Neuro and the RICU team.

For reference, I’m reprinting here the American Heart Association’s Guidelines for Early Management of Pts with Ischemic Stroke (2003) which is all about high blood pressure, and states:

Because of these conflicting issues and the lack of unambiguous data, the appropriate treatment of the blood pressure in the setting of acute ischemic stroke remains controversial. In a majority of patients, a decline in blood pressure without any specific medical treatment will occur.^60,140 The blood pressure often falls spontaneously when the patient is moved to a quiet room, the bladder is emptied, pain is controlled, and the patient is allowed to rest. In addition, treatment of increased intracranial pressure can result in a decline in arterial blood pressure.

Although there are no definitive data from controlled clinical trials, in the absence of other organ dysfunction necessitating rapid reduction in blood pressure, or in the setting of thrombolytic therapy, there is little scientific basis and no clinically proven benefit for lowering blood pressure among patients with acute ischemic stroke.¹⁴³ In most circumstances, the blood pressure should generally not be lowered. Situations that might require urgent antihypertensive therapy include hypertensive encephalopathy, aortic dissection, acute renal failure, acute pulmonary edema, or acute myocardial infarction.¹⁴⁴

Although severe hypertension might be considered as an indication for treatment, there are no data to define the levels of arterial hypertension that mandate emergent management.¹⁴³ The consensus is that antihypertensive agents should be withheld unless the diastolic blood pressure is >120 mm Hg or unless the systolic blood pressure is >220 mm Hg (level V).

When treatment is indicated, lowering the blood pressure should be done cautiously. Parenteral agents such as labetalol that are easily titrated and that have minimal vasodilatory effects on cerebral blood vessels are preferred. In some cases, an intravenous infusion of sodium nitroprusside may be necessary for adequate blood pressure control…. Thrombolytic therapy is not given to patients who have a systolic blood pressure >185 mm Hg or a diastolic blood pressure >110 mm Hg at the time of treatment.

More useful, practical tips on BP control in acute stroke (and its subsequent effects on your decision to thrombolyze) can be found on emcrit.org. As usual, more thoughts on these papers or the topics they raise is welcome in the comments section below.

Posted on Thursday, November 16th, 2006 at 7:16 am by Nick. Filed under Stroke / TIA, Risk Stratification, Journal Club.
You may post a comment.