DOI 10.1212/01.wnl.0000219628.78513.b5 2006;66;1878-1881 Neurology L. G. Stead, R. M. Gilmore, K. C. Vedula, et al. outcome Impact of acute blood pressure variability on ischemic stroke This information is current as of June 26, 2006 http://www.neurology.org/content/66/12/1878.full.html located on the World Wide Web at: The online version of this article, along with updated information and services, is 0028-3878. Online ISSN: 1526-632X. since 1951, it is now a weekly with 48 issues per year. Copyright . All rights reserved. Print ISSN: ® is the official journal of the American Academy of Neurology. Published continuouslyNeurology http://www.neurology.org/content/66/12/1878.full.html Impact of acute blood pressure variability on ischemic stroke outcome L.G. Stead, MD; R.M. Gilmore, MRCPI; K.C. Vedula, BS; A.L. Weaver, MS; W.W. Decker, MD; and R.D. Brown, Jr., MD, MPH Abstract—Background: The authors previously reported a low initial emergency department (ED) blood pressure (BP) to be associated with a significantly increased risk of death at 90 days. In this article, they examine the impact of acute BP variability following onset of ischemic stroke. Methods: The study cohort consisted of 71 patients with stroke onset less than 24 hours. BP measurements were obtained every 5 minutes for the duration of the patient’s ED stay. During the first 180 minutes, the median number of readings per patient was 9 (range 2 to 30). Results: The baseline median systolic (sBP) and diastolic (dBP) BPs were not different for the patients that died within 90 days compared with those that were alive (p � 0.91 for sBP and p � 0.27 for dBP). Patients who died within 90 days had a greater differential in their dBP during the first 180 minutes than the patients that were alive after 90 days (median 44.5 vs 25 mm Hg; Wilcoxon rank sum test, p � 0.001). A similar result was observed for sBP (median 47 vs 30 mm Hg; p � 0.047). Conclusion: Wide fluctuation of blood pressure in the first 3 hours of the emergency department stay in patients with acute ischemic stroke appears to be associated with an increased risk of death at 90 days. NEUROLOGY 2006;66:1878–1881 Previous data demonstrate the detrimental effect of both low and high mean arterial pressures in pa- tients presenting with acute ischemic stroke (AIS) to the emergency department (ED), using 90-day mor- tality as the primary endpoint. These previous stud- ies were based on a single blood pressure (BP) measurement performed at triage.1 In the current study, we evaluated whether changes in systolic (sBP) or diastolic (dBP) BP during the course of the ED are associated with 90-day mortality in AIS. Methods. This study was conducted at Saint Mary’s Hospital, a tertiary care academic medical center, affiliated with Mayo Clinic in Rochester, MN. The annual ED census is approximately 70,000. The initial study population consisted of 480 consecutive patients who presented to the ED with AIS within 24 hours of symptom onset between mid-December 2001 and March 2004. For purposes of obtaining consistent follow-up, this sample was further limited to the 381 patients who resided in the local county or the sur- rounding nine-county area. All patients provided research autho- rization. The initial BP was available for 357 of these 381 patients. As part of our ED’s acute stroke protocol, all patients undergo continuous BP monitoring while in the ED. The Philips M3046A Patient Monitoring System (Philips Medical Systems, Andover, MA) is used to collect sBP and dBP automatically every 5 minutes, which are then recorded electronically and used to generate a trend report. At the time of this study, BP trend re- ports were available on 71 patients. The following additional patient information was abstracted from the medical record: date of birth, gender, date of ED presen- tation, date of death or last follow-up, and medical history infor- mation such as history of hypertension, diabetes, valvular heart disease, and atrial fibrillation. Patients were defined as having hypertension if they self reported a history of hypertension, were on antihypertensive therapy, or had a history of sBP �160 mm Hg or dBP �90 mm Hg on two separate readings during prior routine office visits. Patient medical records were also reviewed for the results of cardiac, vascular, and brain imaging studies and as- signed a stroke subtype in accordance with the classification de- scribed in the Trial of ORG 10172 in Acute Stroke Treatment (TOAST).2 Follow-up was updated for the final data analysis using the date of the last service or dismissal available from the regis- tration databases. In addition, dates and causes of death were ascertained from the State of Minnesota Electronic Death Certifi- cate Data. The NIH Stroke Scale (NIHSS) score was calculated by a physician certified in the NIHSS, based on previously validated methods.3,4 The scoring was derived from documentation of the neurologic examination performed by a neurologist at the time of ED presentation. This study was approved by the Mayo Founda- tion Institutional Review Board. Characteristics of the patients with available trend reports (n � 71) were compared with the characteristics of the rest of the patients presenting with acute ischemic stroke in the study sam- ple (n � 357 � 71 � 286) to evaluate for potential patient inclu- sion bias. Comparisons between these two groups were performed using the �2 test, two-sample t test, Wilcoxon rank sum test, as appropriate. Among the sample of patients with available trend reports, the primary endpoint was 90-day mortality. As all of these patients had at least 90 days of follow-up, the endpoint was evaluated as a simple binary endpoint rather than using time-to-event methodol- ogy that takes into account varying lengths of follow-up. Using the readings recorded during the first 180 minutes, we determined the following values for both sBP and dBP: baseline (average of values recorded at time 0 and 10 minutes), maximum differential (defined as maximum – minimum), percentage change of the final reading from baseline, and slope (derived from fitting a linear regression line to the measures over time). For each of these measurements, the Wilcoxon rank sum test was used to evaluate whether the median value was significantly different between pa- tients that did vs did not die within 90 days. In addition, two logistic regression models were fit to evaluate the association be- tween 90-day mortality and log (maximum differential), sepa- rately for sBP and dBP. Each model was adjusted for age and log (NIHSS). For maximum differential and NIHSS, a logarithmic From the Department of Emergency Medicine (L.G.S., R.M.G., W.W.D.), Division of Biostatistics (A.L.W.), and Department of Neurology (R.D.B.), Mayo Clinic, Rochester, and Department of Biomedical Engineering (K.C.V.), University of Minnesota, Minneapolis, MN. Disclosure: The authors report no conflicts of interest. Received August 23, 2005. Accepted in final form March 13, 2006. Address correspondence and reprint requests to Dr. L.G. Stead, Department of Emergency Medicine, Mayo Clinic, 200 First St. SW, Generose G-410, Rochester, MN 55905; e-mail:
[email protected] 1878 Copyright © 2006 by AAN Enterprises, Inc. transformation was applied to the values to obtain distributions that were more normally distributed based on visual inspection of histograms and normal probability plots. All calculated p values were two sided, and p values less than 0.05 were considered sig- nificant. Statistical analyses were performed using the SAS soft- ware package (SAS Institute, Cary, NC). Results. Among the 71 patients in the cohort, 36 (51%) were men; the mean age was 75.0 years (SD 14.7 years; range 26 to 98 years). Approximately one-fifth of the cohort had coronary artery disease requiring coronary artery by- pass grafting. Forty-two percent of the cohort had a history of atrial fibrillation. One-half of the cohort had a history of hyperlipidemia and one-fourth a history of prior stroke. Fourteen percent were current smokers. The racial compo- sition of the cohort was as follows: 85% Caucasian, 7% Asian, 1% Hispanic, 6% chose not to identify themselves, and 1% identified themselves as “other.” All patients presented within 24 hours of stroke symp- tom onset (figure 1), with the majority (82%) presenting within 2 hours of symptom onset. The mean time to ED presentation from stroke ictus was 5.26 hours with an interquartile range of 1.51 to 7.6 hours. Seventy-five per- cent of the cohort had a history of hypertension. None of the patients received antihypertensive therapy in the ED. The NIHSS score ranged from 1 to 37 with median of 7 (mean [SD] 11.8 [10.4]). Based on the TOAST classification for stroke subtype, 32 (45%) of the patients had cardioem- bolic strokes, 10 (14%) had large vessel disease, and 6 (8%) had small vessel disease. The average (SD) dBP and sBP at baseline were 81.9 (20.7) and 169.3 (34.6) mm Hg. A total of 20 patients were dead at 90 days, yielding an overall 90-day mortality of 28%. The most common cause of death by far was the acute stroke itself. Other causes included respiratory distress/pneumonia, cardiovascular (myocardial infarction and congestive heart failure), and renal failure. The characteristics of the 71 patients with available trend reports were compared with those of the remaining patients in the sample without available trend reports. There was no significant difference between the two sub- sets of patients in terms of gender, history of hypertension, age, and initial dBP and sBP. However, patients in the subset with available trend reports had a higher median NIHSS scores (median 7 vs 5; p � 0.004), were more likely to have had cardioembolic strokes (45 vs 28%; p � 0.005), and tended to be more likely to die within 90 days (28 vs 19%; p � 0.098). During the first 180 minutes, the median number of readings per patient was 9 (range 2 to 30). The baseline median sBP and dBP were not different for the patients that died within 90 days compared with those that were alive (p � 0.91 for sBP and p � 0.27 for dBP; table). Patients who died within 90 days had a greater differential in their dBP during the first 180 minutes than the patients that were alive after 90 days (median 44.5 vs 25 mm Hg; Wilcoxon rank sum test, p � 0.001). A similar result was observed for sBP (median 47 vs 30 mm Hg; p � 0.047). The distribution of the dBP differentials for the alive and de- ceased groups is summarized in figures 2 and 3. As demise after stroke is clearly related to age and severity, we fit a logistic regression model to evaluate the association between 90-day mortality and BP differential adjusting for these two factors. Data on stroke severity were obtained in the form of the NIHSS. Within this sample, the patients that died within 90 days had a higher NIHSS score (median 24.5 vs 5.0; p � 0.001) and tended to be older (mean 78.6 vs 73.6; p � 0.19). How- ever, women were not identified as more likely that men to die within 90 days (31 vs 25%; p � 0.55). For dBP, the association between the maximum differential and 90-day mortality remained significant in the model adjusted for age Figure 1. Time to emergency department presentation from symptom onset. Table Comparison of blood pressure values derived from measurements obtained during the first 180 minutes, between patients with vs without 90-day mortality 90-day mortality Yes, n � 20 No, n � 51 Wilcoxon rank sum test p value Systolic BP, mm Hg Baseline 0.91 Mean (SD) 171.1 (31.8) 168.6 (35.9) Median (IQR) 174.8 (146, 188) 160 (143, 194) Differential* 0.047 Mean (SD) 50.8 (28.8) 37 (23.0) Median (IQR) 47 (29, 70) 30 (22, 50) Slope, change in mm Hg/10 min 0.72 Mean (SD) �1.3 (2.6) �0.9 (3.2) Median (IQR) �0.1 (�2.4, �0.5) �0.5 (�1.9, �0.7) % change from baseline† 0.58 Mean (SD) 1.7 (17.7) �2.7 (12.6) Median (IQR) �0.6 (�12.0, �9.6) �2.6 (�11.3, 6.1) Diastolic BP, mm Hg Baseline 0.27 Mean (SD) 86.4 (25.4) 80.2 (18.6) Median (IQR) 90.5 (67, 100) 80 (67,90) Differential* �0.001 Mean (SD) 46 (20.7) 27.3 (17.1) Median (IQR) 44.5 (31,58) 25 (16, 37) Slope, change in mm Hg/10 min 0.61 Mean (SD) �1.2 (2.7) �0.7 (2.1) Median (IQR) �0.5 (�2.2, �0.6) �0.2 (�1.3, �0.7) % change from baseline† 0.73 Mean (SD) �9.2 (24.4) �4.3 (15.7) Median (IQR) �7.4 (�18.3, �9.5) �0.6 (�14.0, �6.1) * Differential, maximum – minimum. † Percentage change from baseline ([final – baseline]/baseline) * 100%. IQR � interquartile range; BP � blood pressure. June (2 of 2) 2006 NEUROLOGY 66 1879 and NIHSS (p � 0.017). The result was not significant for the sBP. Discussion. We found that fluctuation in BP early in the course of ischemic stroke is associated with poor 90-day survival. This effect is independent of prognostic factors such as stroke severity, age, and gender. All our recordings were done within 24 hours of stroke onset during the patients’ initial presenta- tion to the ED, with over 80% done within 2 hours of symptom onset. The direction of the change, that is, whether an increase or a decrease in BP, did not matter. Rather, it was the variability, particularly in dBP, during the course of the ED stay that was most predictive of poor outcome. Previous work has al- ready documented that early hypotension or hyper- tension is associated with increased mortality.1 The current data show that large or extreme variations in dBP and sBP pressure in the acute period are also predictive of a poor outcome. Autonomic dysfunction in ischemic stroke has been extensively investigated.5-7 Some have investi- gated the diurnal BP variability 24 to 72 hours after stroke onset.8,9 Others, as in the current study, ex- amined the course of BP earlier in the course of acute stroke, but data were collected within 72 hours of ictus rather than within 24 hours as in the current study. One study demonstrated that beat-to-beat BP variability was increased after AIS, whereas carotid baroreceptor sensitivity was decreased.10 However, no correlation to clinical outcome was made. Another study, also done within 72 hours of ictus, obtained 10 minute continuous BP recordings, and found that beat-to-beat variability in the mean arterial pressure as well as mean arterial pressure variability was significantly associated with a poor outcome, defined as death or dependency.11 The human brain has a decreased ability to auto- regulate following AIS.12,13 Hence, minor fluctuations in BP may lead to under- or overperfusion of the delicate ischemic neurons. Increases in systemic BP in this setting can lead to increased risk of edema and hemorrhagic transformation, whereas decreases in BP can lead to infarct extension. We believe that the increased mortality observed in this study in those with fluctuating BPs can be attributed to im- paired autoregulation. We postulate that those pa- tients who kept a more constant BP throughout their ED stay had better perfusion of the ischemic penum- bra and hence better outcomes. There are important potential limitations in the current study. First, the study had relatively small numbers due to continuous BP reports not being available on all patients. In addition, given that there were 20 deaths, the multivariate modeling was limited according to the rule of 10 of allowing 1 inde- pendent variable (specifically 1 df) in the model per every 10 events. Whereas the baseline characteris- tics were similar in this group when compared with our overall stroke registry cohort of all consecutive AIS patients presenting to the ED, the current study’s cohort had higher initial NIHSS, higher inci- dence of death at 90 days, and a higher incidence of cardioembolic strokes. This implies that this current subset of patients had more severe strokes, and therefore these data are not necessarily generaliz- able to all acute stroke subjects. This study is a hypothesis-generating pilot, and these data need to be replicated in a larger cohort, preferably in con- junction with autonomic testing to document im- paired control of BP and heart rate. References 1. Stead LG, Gilmore RM, Decker WW, Weaver AL, Brown RD Jr. Initial emergency department blood pressure as a predictor of survival after acute ischemic stroke. Neurology 2005;65:1179–1183. 2. Adams HP, Bendixen BH, Kappelle LJ, et al. Classification of subtype of acute ischemic stroke: definitions used in a multicenter trial. Stroke 1993;24:35–41. Figure 2. Summary of the differential (max � min) in dia- stolic blood pressure in the first 180 minutes. Figure 3. Summary of the differential (max � min) in sys- tolic blood pressure in the first 180 minutes. 1880 NEUROLOGY 66 June (2 of 2) 2006 3. Kasner SE, Chalela JA, Luciano JM, et al. Reliability and validity of estimating the NIH Stroke Scale score from medical records. Stroke 1999;30:15347. 4. Williams LS, Yilmaz EY, Lopez-Yunez AM. Retrospective assessment of initial stroke severity with the NIH Stroke Scale. Stroke 2000;31:858– 862. 5. Robinson TG, James M, Youde J, et al. Cardiac baroreceptor sensitivity is impaired after acute stroke. Stroke 1997;28:1671–1676. 6. Robinson T, Potter J. Cardiopulmonary and arterial baroreflex- mediated control of forearm vasomotor tone is impaired after acute stroke. Stroke 1997;28:2357–2362. 7. Nishioka Y, Sashika H, Andho N, Tochikubo O. Relation between 24-h heart rate variability and blood pressure fluctuation during exercise in stroke patients. Circ J 2005Jun;69:717–721. 8. Jain S, Namboodri KK, Kumari S, Prabhakar S. Loss of circadian rhythm of blood pressure following acute stroke. BMC Neurol 2004;4:1. 9. Hickey JV, Salmeron ET, Lai JM. Twenty-four-hour blood pressure variability after acute ischemic stroke. Crit Care Nurs Q 2002;25:1– 12. 10. Dawson SL, Panerai RB, Potter JF. Dynamic cerebral autoregulation and beat to beat blood pressure control are impaired in acute ischaemic stroke. J Neurol Neurosurg Psychiatry 2002;72:467–472. 11. Dawson SL, Manktelow BN, Robinson TG, Panerai RB, Potter JF. Which parameters of beat-to-beat blood pressure and variability best predict early outcome after acute ischemic stroke? Stroke 2000;31:463– 468. 12. Lassen N. The luxury perfusion of the brain and its possible relation to acute metabolic acidosis localized in the brain. Lancet 1966;113– 115. 13. Fieschi C, Agnoli A, Battistini N, Bazzao L, Prencipe M. Derangement of regional cerebral blood flow and its regulatory mechanisms in acute cerebrovascular lesions. Neurology 1968;18:1166–1179. MARK YOUR CALENDARS! Plan to attend the 59th Annual Meeting in Boston, April 28–May 5, 2007 The 59th Annual Meeting Scientific Program highlights leading research on the most critical issues facing neurolo- gists. More than 1,000 poster and platform presentations cover the spectrum of neurology—from updates on the latest diagnostic and treatment techniques to prevention and practice management strategies. For more information contact AAN Member Services at
[email protected]; (800) 879-1960, or (651) 695- 2717 (international). AAN 60th Annual Meeting in Chicago, Illinois April 12–19, 2008 Chicago, IL AAN 61st Annual Meeting in Seattle, Washington April 25–May 2, 2009 June (2 of 2) 2006 NEUROLOGY 66 1881 DOI 10.1212/01.wnl.0000219628.78513.b5 2006;66;1878-1881 Neurology L. G. Stead, R. M. Gilmore, K. C. Vedula, et al. Impact of acute blood pressure variability on ischemic stroke outcome This information is current as of June 26, 2006 Services Updated Information & http://www.neurology.org/content/66/12/1878.full.html including high resolution figures, can be found at: References 1 http://www.neurology.org/content/66/12/1878.full.html##ref-list- at: This article cites 11 articles, 9 of which you can access for free Citations icles http://www.neurology.org/content/66/12/1878.full.html##otherart This article has been cited by 9 HighWire-hosted articles: Subspecialty Collections http://www.neurology.org//cgi/collection/prognosis Prognosis http://www.neurology.org//cgi/collection/infarction Infarction ease_stroke http://www.neurology.org//cgi/collection/all_cerebrovascular_dis All Cerebrovascular disease/Stroke following collection(s): This article, along with others on similar topics, appears in the Permissions & Licensing http://www.neurology.org/misc/about.xhtml#permissions or in its entirety can be found online at: Information about reproducing this article in parts (figures,tables) Reprints http://www.neurology.org/misc/addir.xhtml#reprintsus Information about ordering reprints can be found online: http://www.neurology.org/content/66/12/1878.full.html http://www.neurology.org/content/66/12/1878.full.html##ref-list-1 http://www.neurology.org/content/66/12/1878.full.html##ref-list-1 http://www.neurology.org/content/66/12/1878.full.html##otherarticles http://www.neurology.org/content/66/12/1878.full.html##otherarticles http://www.neurology.org//cgi/collection/all_cerebrovascular_disease_stroke http://www.neurology.org//cgi/collection/all_cerebrovascular_disease_stroke http://www.neurology.org//cgi/collection/infarction http://www.neurology.org//cgi/collection/prognosis http://www.neurology.org/misc/about.xhtml#permissions http://www.neurology.org/misc/addir.xhtml#reprintsus