Racial Inequity of Access to Carotid Imaging
To the Editor:
While we welcome research into stroke in patients of African and Caribbean origin, we have a number of concerns regarding the conclusions of Oddone et al on racial differences in the use of carotid imaging.1
Lacunar stroke may give a presentation identical to that of a transient ischemic attack (TIA). There is a high prevalence of lacunar stroke in people of Afro-Caribbean heritage both in the United States and the United Kingdom. Furthermore, there is a negative correlation between lacunar stroke and carotid disease.2 Thus, when the strictest indication for carotid imaging is used (to define a symptomatic stenosis in a patient suitable for surgery), a higher proportion of African-American patients may not have qualified. (The brain imaging was not reviewed in the study.)
Of the African-Americans in the TIA group, 34.9% (30 of 86) had had a previous TIA and 33.7% (29 of 86) a previous stroke, compared with 15.2% (28 of 184) and 16.3% (30 of 184) of white patients. As the study only analyzed photocopies of case records from the index admission and for 6 months afterwards, it is conceivable that some of these patients’ carotids had been imaged previously. If a normal carotid study had been performed recently, repeat imaging might be considered an inappropriate use of resources. If this situation held for even 4 African-American patients, the statistical significance for the statement “fewer black than white patients with TIA…received…studies of their carotid arteries” is lost.
In terms of the lower proportion of African-Americans who underwent surgery for appropriate and uncertain indications, Oddone’s own work has shown differences in risk acceptance between different ethnic groups.3 It is unclear how many patients were offered an endarterectomy and refused.
The implication that physicians deny appropriate investigation and treatment to a particular ethnic group is a serious one. While there may be many reasons, including genetics, risk factors, vascular pathophysiology and sociocultural factors, for the excess of stroke in people of African and Caribbean heritage, we do not feel that inequity of access to carotid imaging has been proved.
- Copyright © 1999 by American Heart Association
Oddone EZ, Horner RD, Sloane R, McIntyre L, Ward A, Whittle J, Passman LJ, Kroupa L, Heaney R. Race, presenting signs and symptoms, use of carotid artery imaging, and appropriateness of carotid endarterectomy. Stroke.. 1999;30:1350–1356.
Mead GE, Wardlaw JM, Lewis SC, McDowall M, Dennis MS. Can clinical features be used to identify patients with severe carotid stenosis on Doppler ultrasound? J Neurol Neurosurg Psychiatry.. 1999;66:16–19.
We appreciate the comments of Drs Evaus and Kaltra concerning our manuscript and agree strongly that investigators need to exercise caution when interpreting findings regarding racial/ethnic differences in utilization of health care. They raise 3 important points. First, they suggest that a racial difference in stroke subtype, particularly lacunar stroke, may account for the lower proportion of black patients who received carotid imaging. We did not assess stroke subtype in this study because, as a medical record review, determination of stroke subtype is unreliable. Also, patients with small-vessel disease may also have carotid artery disease. If it were possible by some means short of carotid imaging to accurately discern which symptomatic black patients were likely to not have carotid stenosis (or whose carotid stenosis is unlikely to lead to future stroke), we could appropriately avoid many imaging studies. Since that is not possible, it may be prudent to perform a noninvasive imaging study of the carotid arteries even in black patients with lacunar infarctions.
Second, Drs Evaus and Kaltra correctly note that we counted imaging during the index hospitalization and for 6 months after discharge, potentially missing patients who had previous carotid studies that did not reveal high-grade stenosis. The issue here is whether old imaging studies from prior symptomatic events are sufficient to rule out new significant disease in a patient with new symptoms. By focusing on imaging studies after the new event, we were assuming that new symptoms typically require new imaging, since carotid disease is progressive. However, it is true that we would have missed some black patients who had had recent negative imaging studies and for whom new tests would not be necessary. We should note that this is unlikely, since such studies would usually be reported in the index hospitalization.
Last, they infer that our report implied that black patients were being denied appropriate referral for carotid endarterectomy. As explained in the discussion, we do not want to imply, nor do we want readers to infer, that this study demonstrates denial of appropriate care. Indeed, as preferences for this procedure may differ by race, black patients may be less likely to accept the procedure if offered. As we discussed, there were too few black patients who received endarterectomy in our study to make inferences concerning the differential use of this procedure by race. However, we do feel that the differential referral for carotid artery imaging, the vast majority of which is noninvasive ultrasound, cannot be explained by this phenomenon. Informed decisions can be made only when all of the important information is in hand. Knowledge of the degree of stenosis is essential for patients who may be eligible for endarterectomy. The main finding of our study is that there was a differential referral for imaging that we cannot otherwise explain. We are currently engaged in a large prospective cohort study designed to determine the degree to which patient and physician factors predict referral. When this study is completed we hope to better understand the specific cause of differential referral should we find it.