a b s t r a c t

Introduction: Retrospective data indicates that dehydration in acute ischemic stroke patients may be common, even though these patients frequently have Elevated blood pressure. We sought to evaluate clinical and labora- tory measures of intravascular volume status compared to more objective measures using ultrasound measure- ments of the Inferior vena cava .

Methods: This was a prospective observation study of acute ischemic stroke patients in the emergency depart- ment. Patients with NIH stroke scale >=4 within 12 h of symptom onset were included. A trained ultrasonographer performed bi-dimensional imaging of the IVC with passive respiration to determine the percent inspiratory col- lapse and maximum diameter. We defined low intravascular volume as N50% IVC collapse and a maximal diam- eter b 2.1 cm. Analysis was limited to patients with confirmed ischemic stroke.

Results: There were 42 patients, of whom 31 had confirmed acute ischemic stroke. The mean age was 65 +- 15

years, 52% were female, and 71% were hypertensive. The median NIH stroke scale score was 7 (IQR 5-15). Based on IVC ultrasound, low intravascular volume was present in 63% (95% CI 44-80%) of patients. A higher pro- portion of Hypertensive patients had low intrasvascular volume (72% vs. 33%). There was poor correlation be- tween IVC assessment of intrasvascular volume and blinded clinician assessment or laboratory markers of dehydration.

Conclusion: The majority of ED acute ischemic stroke patients in this sample were hypertensive and demonstrat- ed low intravascular volume based on IVC ultrasound.

(C) 2018

Introduction

Dehydration or reduced intravascular volume has been associated with deleterious effects in acute ischemic stroke. In one study, as many as 60% of stroke patients had laboratory evidence of dehydration [1]. This finding was associated with poorer functional recovery. Other studies have similarly found an association between dehydration and poor outcomes [2-4]. A key limitation of these studies, however, has been reliance on indirect laboratory evidence to define dehydration. Prior studies have used elevated Blood urea nitrogen or high serum osmolality as markers of low intrasvascular volume. Further complicating the clinical picture is that as approximately 75% of acute is- chemic stroke patients have elevated Blood pressure on presenta- tion [5]. It can be challenging to identify such patients as having low intravascular volume. Missing such patients, however, can be deleteri- ous. Candidates for endovascular or thrombolytic therapies frequently

* Corresponding author at: Henry Ford Hospital, Department of Emergency Medicine, 2799 W Grand Blvd, Detroit, MI 48202, United States

E-mail address: [email protected] (J.B. Miller).

https://doi.org/10.1016/j.ajem.2018.01.071

0735-6757/(C) 2018

receive rapid acting antihypertensive agents that could overshoot treat- ment goals and reduce cerebral perfusion when intravascular volume is low.

Ultrasonography provides more direct, non-invasive measurements of volume status. Measurements of small inferior vena cava diam- eter and greater respiratory collapse indicate low intravascular volume [6-8]. A bedside test to identify acute ischemic stroke patients that have low intravascular volume could be useful in their early management. Hence, the primary aim of this study was to assess the rate of low intra- vascular volume in acute stroke patients using IVC measurements com- pared to clinical assessment and BP. We secondarily sought to compare these measurements to Laboratory markers of dehydration.

Methods

This was a prospective observation study of acute ischemic stroke patients within 12 h of symptom onset. The setting was a high volume, urban emergency department within a primary stroke center. Patients were eligible if they had suspected ischemic stroke and a National Insti- tutes of Health Stroke Scale (NIHSS) score of 4 or greater. Exclusion

J.B. Miller et al. / American Journal of Emergency Medicine 36 (2018) 1018-1021 1019

criteria were known hemorrhagic stroke, age b18, and mechanical ven- tilation. Institutional review board approved the study and granted waiver of informed consent.

Patients were enrolled in a non-consecutive manner based on study team availability. Study team members that performed ultra- sound image acquisition received formal training in IVC imaging and were required to demonstrate proficiency after 10 practice scans. A fellowship trained ultrasound staff oversaw training. Image acquisition was obtained via the subxiphoid window with the patient supine in bed and the head of the bed at 0-15?. Investiga- tors used a low-frequency, Phased-array probe. Transverse and sag- ittal images were stored and measured in b-mode to determine maximum and minimum IVC diameter with respiration. The mea- surements were taken just caudal to the first hepatic vein, within two centimeters of the atriocaval junction.

Investigators collected laboratory, presenting BP, demographic and clinical stroke characteristics for each patient. We chose presenting BP as has previously been used to reflect the acute hypertensive response common in stroke [9]. Laboratory data included blood urea nitrogen (BUN), creatinine (Cr), and the calculated, expected serum osmolality. In addition, investigators collected a standardized query from each primary clinical physician. This query asked clinicians to provide their clinical assessment regarding each patient’s Intravascular volume status, based on his or her initial physical examination (including BP) and history of the patient. Investigators blinded the treating team to the ultrasound images and measurements. Investigators followed patients through their hospital stay. The final diagnosis of acute ischemic stroke versus an alternative diagnosis was determined by magnetic resonance imaging and the assessment of the treating vascular neurologist.

The primary outcome was the rate of low intravascular volume using IVC measurements in relation to BP and clinical assessment. We classified patients as hypertensive if their presenting systolic BP (SBP) was N140 mm Hg. Following guidelines from the American Society of Echocardiography, we used two measurements to define intravascular volume: caval collapse and maximal IVC diameter [9]. Collapse (Caval index) was measured as the maximum IVC diameter minus the minimum diameter, divided by the maximum diameter and multiplied by 100%. We considered a patient to have low intravascular volume if the maximal IVC diameter was b2.1 cm and the caval index was N50%. These findings indicate a central venous pressure b 5 mm Hg [10-12]. Patients with a maximal IVC diameter >= 2.1 cm but caval collapse N50% through the respiratory cycle were considered to have intermediate volume status. So too were patients with a maximal diameter b 2.1 cm and b50% collapse. Intermediate volume status is associated with central venous pressure of 5-10 mm Hg [10]. Finally, patients with a caval index <=50% and a maximal diameter >= 2.1 cm were estimated to have high intrasvascular volume.

Secondarily, we tested the correlation between IVC determina-

tion of intravascular volume status and laboratory markers. The spe- cific laboratory markers were a BUN/Cr ratio N 20 or a calculated serum osmolality above the normal reference range (285-295 mOsm/kg). We calculated the sample size for this study using estimates for a one-sample, binomially distributed outcome. We based sample size on a null hypothesis probability of 30% and alternative hypothesis probability of 50% of patients having low intravascular volume. Hence, we estimated a sample size of 39 subjects to have 80% power with 0.05 alpha. Statistical analysis consisted of Cohen’s kappa coefficient to compare agreement between measures. Analysis included chi-square or Fisher’s exact test to compare categorical variables between patients found to have low intravascular volume or not. We used a student’s t-test or Wilcoxon rank sum test depending on the normality of the data to compare continuous variables for this portion of analysis. All analysis was performed in SAS 9.4 (Cary, NC).

Results

We performed IVC imaging on 42 patients, 31 (74%) of which had a confirmed diagnosis of acute ischemic stroke. Of the confirmed stroke patients, 20 (65%) had large vessel, anterior circulation stroke. There were 8 (26%) patients with lacunar, 2 with posterior stroke, and 1 with a mixed distribution stroke. Of the 11 patients without ischemic stroke, 5 had transient ischemic attack, 5 stroke mimics, and 1 hemor- rhagic stroke.

In the primary analysis, we included the 31 patients with confirmed acute ischemic stroke. This cohort of 31 confirmed ischemic stroke pa- tients was 52% female with a mean age of 65 +- 15 years. Most patients were African American (77%) and 16% were Caucasian. Most were also hypertensive upon arrival (71%). The median NIHSS score was 7 (IQR 5-15). Mean presenting SBP was 164 +- 31 mm Hg and mean calculated serum osmolality was 295 +- 14 mOsm/kg. The mean BUN/Cr ratio was

14.1 +- 6.6. The proportion of patients with elevated serum osmolality was 16% and with elevated BUN/Cr ratio 10%. Table 1 demonstrates the presenting demographic and clinical characteristics.

The median time from stroke symptoms onset to study assessment was 204 min (IQR 74-720). The median imaging time for sonographic assessment was 5 (IQR 2-10) min. Table 2 shows the IVC measurements and agreement with other measures of volume status. Mean maximal IVC diameter was 1.5 +- 0.48 cm and mean caval collapse was 57 +- 32%. The proportion of patients with low intravascular volume, defined as a maximal IVC diameter b 2.1 cm and collapse N50%, was 63% (95% CI 44-80%). There were 2 (7%) patients with high and 9 (30%) with inter- mediate intravascular volume. Patients that were hypertensive on pre- sentation had higher rates of low intravascular volume (72%) compared to normotensive patients (33%), though this did not reach statistical significance (p = 0.056).

Clinicians estimated that 8 (27%) of patients had low intravascular

volume. There was poor agreement between the IVC measurement of volume status and blinded clinical assessment (kappa -0.36, 95% CI

-0.68 to -0.05). Furthermore, when a hypertensive patient had low intravascular volume by IVC measurement, clinical assessment agreed with volume status in only 6% of cases compared to 50% of cases when the patient was normotensive (p = 0.01). Agreement was also poor be- tween IVC determination of volume status and laboratory markers (kappa 0.1, 95% CI -0.1 to 0.3). Univariate comparison between

Table 1

Demographic and clinical characteristics of patients with acute ischemic stroke.

Age, mean (SD) 65 (15) years

Female, n (%) 16 (52)

Race, n (%)

Black 24 (77)

Caucasian 5 (16)

Other 2 (7)

NIHSS^a, median (IQR) 7 (5-15)

Time from symptom onset, median (IQR) 204 (74-720) min Comorbidities, n (%)

Hypertension 27 (87)

Diabetes 18 (58)

Coronary artery disease 6 (19)

Atrial fibrillation 5 (16)

Prior stroke 9 (29)

Stroke sub-type

Large vessel, anterior circulation 20 (65)

Lacunar 8 (26)

Posterior circulation 2 (6)

Laboratory results

Serum osmolality, mean (SD) 295 +- 14 mOsm/kg

High serum osmolality, n (%) 5 (16)

BUN/Cr ratio, mean (SD) 14 (7)

BUN/Cr ratio N 20, n (%) 3 (10)

SBP, mean (SD) 159 +- 30 mm Hg

^a NIHSS, National Institutes of Health Stroke Scale; IQR, interquartile range; SBP, systolic blood pressure; osmolality is calculated serum osmolality.

1020 J.B. Miller et al. / American Journal of Emergency Medicine 36 (2018) 1018-1021

Table 2

Caval measurements compared to other measures of intravascular volume.

Imaging time, median (IQR)^a 5 (2-10) min

Clinician assessment, n (%)

Patient with normal to high intravascular volume 22 (73) Patient with low intravascular volume 8 (27)

IVC measurements

Maximal, mean (SD) 1.5 (0.5) cm

Minimal, mean (SD) 0.7 (0.6) cm

Maximal b 2.1 cm, n (%) 28 (90)

Percent collapse, mean (SD) 57 +- 32%

IVC defined intravascular volume, n (%)

Low 19 (63)

Indeterminate 9 (30)

High 2 (7)

Agreement between measures, kappa (95% CI)

Clinician assessment and IVC -0.35 (-0.67 to -0.05)

BUN/Cr N 20 and IVC b 2.1 cm 0.11 (-0.09 to 0.31)

High serum osmolality and IVC b 2.1 cm 0.11 (-0.09 to 0.31)

^a IQR, interquartile range; IVC, inferior vena cava; CI, confidence interval; BUN/Cr, blood urea nitrogen to creatinine ratio.

patients defined as having low versus intermediate or high-volume sta- tus was limited by the low number of patients in the latter category. Nonetheless, females with acute stroke were significantly more likely to have IVC measurements consistent with low intravascular volume than males (88% vs. 50%, p = 0.003). Other comparisons of age, race, time since onset of symptoms, NIHSS, presenting SBP, serum osmolality, hematocrit, or BUN/Cr ratio were not statistically significant.

Discussion

This is the first published study we are aware of to investigate the use of IVC ultrasound to assess volume status in acute ischemic stroke patients. The results support the hypothesis that low intravascular vol- ume by ultrasonography is common in patients with acute ischemic stroke. The rate of high intravascular volume was b10%. Furthermore, low intravascular volume was common even when patients presented with elevated BP.

The finding of low intravascular volume status being common is consistent with the limited published stroke literature in this area [2,4]. However, laboratory surrogates of intravascular volume status, which are the basis of this literature, correlated poorly with IVC mea- surements in the current study’s cohort. The existing critical care litera- ture shows that physical examination has poor diagnostic accuracy for determining intravascular volume status [13]. In acute ischemic stroke, patients are commonly hypertensive, and clinicians may not readily consider low intravascular volume in their acute evaluation and man- agement [5]. Nonetheless, based on existing data linking low intravas- cular volume to poor outcomes, knowledge of acute ischemic stroke patients’ volume status may be important [1-3]. Data from Schrock, for instance, found that patients with evidence of dehydration had N2 times the odds of death or poor functional outcome at 30 days com- pared to those without dehydration [2]. We did not follow patients in our study to determine functional outcomes.

Current guidelines for acute ischemic stroke management acknowl- edge that intravascular Volume expansion has a potential role in emer- gent management but do not routinely recommended it [14]. Assessment of intravascular volume status by IVC measurements iden- tifies patients that may benefit from volume expansion. A small maxi- mal IVC diameter with high collapsibility identifies patients with low cardiac filling pressure. Volume expansion is unlikely to produce harm and may benefit these patients. Patients estimated to have intermediate intravascular volume may also potentially benefit from volume expan- sion. A plethoric IVC with minimal collapse identifies patients with the potential risk of harm from volume expansion. This latter point can have flaws, however. Patients with chronically elevated right heart vol- ume may still see improvements in cardiovascular hemodynamic

performance with volume expansion. Measurements of caval collapse have approximately 84% sensitivity for determining volume responsive- ness [8]. Patients with plethoric IVC measurements may require more advanced measurements of hemodynamic change such as stroke vol- ume to determine potential benefit from volume expansion.

Notable limitations of this study include the relatively small sample

size and convenience sampling methods that may introduce selection bias. We did not use gold standard evaluation of volume responsiveness using invasive hemodynamic monitoring. Such measurements are im- practical in the emergency management of acute stroke patients. This was study was done at a single institution. We did not assess reproduc- ibility of image acquisition. Nevertheless, our image acquisition time was brief (median time 5 min), and we were able to obtain adequate images in all enrolled patients. As described above, IVC diameter can be influenced by several conditions including Right heart failure and pul- monary hypertension. In patients with such comorbidities, IVC ultra- sound may produce a false negative test for Volume depletion. Age may also influence IVC diameter such that older patients have smaller IVC diameter [15]. Finally, we acknowledge that patients defined as hy- pertensive based on their presenting BP may have had spontaneous BP decline. We did not capture subsequent BP measurements to determine how frequently this occurred.

Conclusion

In this sample of acute ischemic stroke patients, a majority were hy- pertensive and found to have ultrasound evidence of low intravascular volume. Clinical assessment identified a minority of these patients. Fur- ther studies are indicated to reproduce these results in diverse popula- tions and to identify whether clinical outcomes are affected by volume expansion in appropriately identified patients.

Presentations

A portion of this work was presented at the American Heart Associ- ation International Stroke Conference, January 2016, Los Angeles, CA

Author contributions

Concept and design: JBM, CAL, JLC; data acquisition, analysis, inter- pretation: JBM, AL, JPS, SM, LR, CAL; manuscript preparation and critical revision: JBM, MT, CAL.

Conflict of interest disclosure

The authors (JBM, AL, JPS, MT, JLC, SM, LR, CAL) report no conflict of interest.

References

1. Rowat A, Graham C, Dennis M. Dehydration in hospital-admitted stroke patients: detection, frequency, and association. Stroke 2012;43(3):857-9.
2. Schrock JW, Glasenapp M, Drogell K. Elevated blood urea nitrogen/creatinine ratio is associated with poor outcome in patients with ischemic stroke. Clin Neurol Neurosurg 2012;114(7):881-4.
3. Bhalla A, Sankaralingam S, Dundas R, Swaminathan R, Wolfe CD, Rudd AG. Influence of raised plasma osmolality on clinical outcome after acute stroke. Stroke 2000; 31(9):2043-8.
4. Rodriguez GJ, Cordina SM, Vazquez G, Suri MF, Kirmani JF, Ezzeddine MA, et al. The hydration influence on the risk of stroke (THIRST) study. Neurocrit Care 2009;10(2): 187-94.
5. AlSibai A, Qureshi AI. Management of acute hypertensive response in patients with ischemic stroke. Neurohospitalist 2016;6(3):122-9.
6. Dipti A, Soucy Z, Surana A, Chandra S. Role of inferior vena cava diameter in assess- ment of volume status: a meta-analysis. Am J Emerg Med 2012;30(8):1414-9 [e1].
7. Muller L, Bobbia X, Toumi M, Louart G, Molinari N, Ragonnet B, et al. Respiratory var- iations of inferior vena cava diameter to predict fluid responsiveness in spontane- ously breathing patients with acute circulatory failure: need for a cautious use. Crit Care 2012;16(5):R188.
8. Preau S, Bortolotti P, Colling D, Dewavrin F, Colas V, Voisin B, et al. Diagnostic accu- racy of the inferior vena cava collapsibility to predict fluid responsiveness in

   J.B. Miller et al. / American Journal of Emergency Medicine 36 (2018) 1018-1021 1021

   spontaneously breathing patients with sepsis and acute circulatory failure. Crit Care Med 2017;45(3):e290-e7.

   Qureshi AI. Acute hypertensive response in patients with stroke: pathophysiology and management. Circulation 2008;118(2):176-87.

9. Rudski LG, Lai WW, Afilalo J, Hua L, Handschumacher MD, Chandrasekaran K, et al. Guidelines for the echocardiographic assessment of the right heart in adults: a re- port from the American Society of Echocardiography endorsed by the European As- sociation of Echocardiography, a registered branch of the European Society of Cardiology, and the Canadian Society of Echocardiography. J Am Soc Echocardiogr 2010;23(7):685-713.
10. Moreno FL, Hagan AD, Holmen JR, Pryor TA, Strickland RD, Castle CH. Evaluation of size and dynamics of the inferior vena cava as an index of right-sided cardiac func- tion. Am J Cardiol 1984;53:579-85.
11. Brennan JM, Blair JE, Goonewardena S, Ronan A, Shah D, Vasaiwala S, et al. Reap- praisal of the use of inferior vena cava for estimating right atrial pressure. J Am Soc Echocardiogr 2007;20:857-61.
12. Bentzer P, Griesdale DE, Boyd J, MacLean K, Sirounis D, Ayas NT. Will this hemody- namically unstable patient respond to a bolus of intravenous fluids? JAMA 2016; 316(12):1298-309.
13. Jauch EC, Saver JL, Adams Jr HP, Bruno A, Connors JJ, Demaerschalk BM, et al. Guide- lines for the early management of patients with acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2013;44(3):870-947.
14. Masugata H, Senda S, Okuyama H, Murao K, Inukai M, Hosomi N, et al. Age-related decrease in inferior vena cava diameter measured with echocardiography. Tohoku J Exp Med 2010;222(2):141-7.