More information about text formats
With interest we read the article by You et al. describing the
identification of risk factors for aneurysm rupture and providing
estimations for the likelihood of rupture in unruptured intracranial
aneurysms . However, we have concerns about the study design, the
presented analysis and the conclusions regarding the predictors for
First, the authors used a study popul...
First, the authors used a study population which included patients
with a newly diagnosed intracranial aneurysm who were treated by surgical
or endovascular techniques. Thus, for the unruptured aneurysms the authors
considered the risk of rupture too high to leave the aneurysm untreated.
Assessing predictors for rupture of aneurysms that are considered having a
risk of rupture high enough to warrant treatment is not very relevant.
From a clinical point of view predicting the risk of rupture of aneurysms
that are not treated (and thus may rupture in the near or distant future)
is much more relevant.
Second, the authors used a nested case control design and matched for
several clinical variables. This is an efficient design for etiologic
research (identification of risk factors), but is not suitable for
prognostic research (determine the risk of rupture based on multiple
predictors). Data from a case control study nested in a cohort of known
size can be used, but only when no matching is performed. The design most
suitable to address prognostic questions is a cohort study in which all
patients with a certain condition are followed for some time to monitor
the development of the outcome .
Finally, and not least important, the underlying frequencies of
potential predictors according to outcome in Table 1 do not correspond
with the univariable odds ratios mentioned in Table 2. For example, we do
not understand how the crude odds ratio for female sex can be different
from 1 (odds ratio 1.4, 95% CI 0.9-2.4) whereas the authors matched on
sex. Also, we do not understand how the crude odds ratio for the
association between hypertension and aneurysm rupture could be 1.9 (95% CI
1.2-3.2) whereas hypertension was less frequent in patients with a
ruptured aneurysm (35%) compared with patients with an unruptured aneurysm
These concerns make interpretation of the presented results extremely
difficult. The study as presented can neither provide reliable information
on predictive risk factors for rupture nor can it form a proper basis for
decisions regarding the optimal therapeutic strategy for unruptured
JP Greving1, Gabriel JE Rinkel2, Ale Algra1,2.
1 Julius Center for Health Sciences and Primary Care, University
Medical Center Utrecht, The Netherlands and 2 Department of Neurology and
Neurosurgery, Rudolf Magnus Institute of Neuroscience, University Medical
Center Utrecht, Utrecht, The Netherlands.
Correspondence to: Dr JP Greving, Julius Center for Health Sciences
and Primary Care, University Medical Center Utrecht, Stratenum 6.131, PO
Box 85500, 3508 GA Utrecht, The Netherlands. email@example.com
Competing interests: None.
 You SH, Kong DS, Kim JS, et al. Characteristic features of unruptured
intracranial aneurysms: Predictive risk factors for aneurysm rupture. J
Neurol Neurosurg Psychiatry 2009; 0: jnnp.2008.169573v1.
 Moons KGM, Royston P, Verhouwe Y et al. Prognosis and prognostic
research: what, why, and how? BMJ 2009;338:b375.