Neurologic Complications of Cerebral Angiography

One might dismiss the article in this issue by Heiserman et al ( 1) on neurologic complications of cerebral angiography by arguing that it brings nothing new to the medical literature. Especially those of us who are quite senior neuroradiologists are familiar with the kinds of findings reported in this paper. The article reports a number of positive and negative correlations of neurologic complications with certain variables , agreeing with some and also disagreeing with some previously reported correlations, again consistent with the lack of complete agreement between earlier publications. However, I am pleased to see this article published in AJNR for a combination of reasons: (a) The subject matter is very important. (b) Scientific knowledge needs to be reevaluated by different investigators and under different circumstances, especially if previous results have been at variance. (c) This is a prospective study involving an unusually large number of studies performed during a relatively short period, considering current patterns of practice. Most previous, similar articles are retrospective studies, and most are from a different era. (d) These are "new times ," with a different mix of patients and generally higher degrees of risk factors among patients referred for cerebral angiography. This article should serve as a timely reminder of some of the problems involved in performing cerebral angiography at present. By coincidence, I was given the opportunity to read the manuscript for this article about 2 weeks after faculty members and fellows in our Division of Neuroradiology had reviewed the complications of cerebral angiographies done by us during the previous year. This was part of a quality control program, which forced us to look more closely, collectively, and more for mally at our results than we have done in the past. Having grown to a much larger section gradually, it now is much more difficult to keep track of how and what our colleagues are doing and how we are performing as a whole in the section. It was not a pleasant experience for me, remembering a different era that predates the experience of most colleagues. I can easily imagine similar reviews going on in other departments, especially given the present medical "climate" of quality control. I am impressed that the authors of the present article performed 1000 cerebral angiographies in only 15 months , no doubt a rather infrequent situation in the United States today. I wonder what the complication rate might be in departments performing only about 100 such examinations yearly, especially if distributed among the same number of neuroradiologists and fellows. Most of all, I wonder what the complication rates may be in departments not reporting and/or unwilling to publish their complication rates , perhaps for fear of scrutiny and criticism by colleagues. I suspect that the present neurologic complication rate of cerebral angiography in many institutions may be significantly higher than most published studies indicate. I promptly reviewed the articles by Mani et a! (2) and Mani and Eisenberg (3, 4), with their analyses of 5000 angiographies from the early computed tomography (CT) era. Without undertaking any formal tabulation of our results and patient demographics some 25 years ago, it was my impression that our experience was similar. At that time, we had two neuroradiology faculty members and two neuroradiology fellows, one in the first and another in the second year of training. The year before we obtained our first CT unit, we performed cerebral angiographies on about 850 patients. Lately, we have had two to three times as many fellows who usually have been closely supervised by three to four times as many faculty members , but