Impact of Surgical Treatment of Unruptured Aneurysms
To the Editor:
In the June 2004 issue of Stroke, Britz et al1 provide us with a clear example of misleading information and nonscientific analysis. The credibility and authority enjoyed by Stroke, combined with a craving for any possible enlightenment on the management of unruptured intracranial aneurysms, strengthen the impact of such a publication and justify the necessity to critically review this work.
I would like to rephrase the problem on which the authors focused:
My hypothesis is that physicians tend to offer a preventive but risky treatment to patients who are most likely to benefit. I believe this hypothesis to be reasonable; in fact, this is all that was shown by the study. If we are to retrospectively compare patients who underwent operation to those who were denied surgery for the same condition, we may end up dividing the population into 2 groups: patients that are likely to survive the intervention and that have a health status conducive to a life expectancy long enough to justify major surgery; and those exposed to higher surgical risk attributed to a more fragile health condition, with a lesser life expectancy.
The method used is a retrospective analysis of demographic and survival data. The results are: Group 2 patients, who were more often younger women (by 10 years), privately insured, with fewer comorbidities, survived at an improved rate as compared with Group 3 patients, older men, publicly insured, with more comorbidities. Since the logic behind this study is circular, the fact that the results confirm the hypothesis is fortunate. The data are empirical but only confirms that in Washington between 1987 and 2001 younger, healthier females would have a longer life expectancy than older, sicker men. But this does not tell us anything about aneurysms.
Somewhat more informative is the high mortality of surgical clipping, as compared with previously published rates,2,3 a common finding when one compares published data with the “real world,” as acknowledged in the text. Given this, it would then seem surprising that the authors conclude that this study supports early surgical management of unruptured aneurysms. The data, invalid from a scientific perspective, collected in a retrospective fashion, with all the potential errors well known to clinical investigators, is interpreted in favor of the biased opinions of the authors: the increased mortality following surgical clipping of unruptured aneurysms, as compared with an estimate of a corresponding population without aneurysms, “must” be related to an unrecognized comorbid state. The alternate hypothesis, that increased mortality is caused by surgery itself, seems so obviously wrong in the eyes of the authors that it is not even mentioned. On the other hand, decreased mortality in the surgical group, as compared with unoperated patients, “must” be credited to surgery. The fact that the 30-day mortality of untreated unruptured aneurysms was extraordinary (7.6%, while annual rupture rates are between 0.05 and 2%2), and higher than the surgical mortality (5.5%), should be sufficient to show the flawed nature of the methodology, and raise the suspicion that the data has nothing to do with aneurysms.
There are many other methodological concerns with this work, such as the inclusion of code ICD-9 437.3 (unruptured aneurysm) as an “event” included in neurological causes of death (another example of circular logic), as well as other codes related to dementia, multiple sclerosis, Parkinson disease and carotid stenosis etc. These are errors committed by the authors, masked by undefined codes, and while they may give indications as to why patients had imaging studies, during which an unruptured aneurysm was found, these inclusions as “neurological-related causes of death”, falsely suggest to readers a causal relationship between mortality and treatment (or not) of the unruptured aneurysm.
Most importantly, works such as this, suffering from systematic errors in study design, comfort our biased opinions and have the potential to contribute to the resistance to clinical trials in this field. We must have the strength to acknowledge our doubts and a strong determination to distinguish desires, beliefs, and facts. There is still no definite scientific evidence to support the surgical or endovascular management of unruptured aneurysms. Nothing short of rigorous scientific methods could give us valid insight to the crucial question: “Should we treat unruptured aneurysms?”4
Britz GW, Salem L, Newell DW, Eskridge J, Flum DR. Impact of surgical clipping on survival in unruptured and ruptured cerebral aneurysms: a population-based study. Stroke. 2004; 35: 1399–1403.
Wiebers DO, Whisnant JP, Huston J 3rd, Meissner I, Brown RD Jr, Piepgras DG, Forbes GS, Thielen K, Nichols D, O’Fallon WM, Peacock J, Jaeger L, Kassell NF, Kongable-Beckman GL, Torner JC. Unruptured intracranial aneurysms: natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet. 2003; 362: 103–110.
Raymond J, Chagnon M, Collet J, Guilbert F, Weill A, Roy D. A randomized trial on the safety and efficacy of endovascular treatment of unruptured intracranial aneurysms is feasible. Interventional Neuroradiology. 2004; 10: 103–112.