The study by Mahr and colleagues, which is reported in this issue of Arthritis & Rheumatism (1), gives us important new insight into the prevalence of Behcet’s disease (Behcet’s syndrome [BS], as we prefer to call this condition) that will be useful in resource allocation and aid in the understanding of disease mechanisms, the traditional aims of epidemiology. In addition, their data allow us to consider how we can improve our diagnostic abilities using epidemiology as a guide. Mahr et al (1) report a BS prevalence rate of 7.1 per 100,000 adults in a suburb of Paris, France. As expected, non-European subgroups, including subjects of Asian and North African descent, had substantially higher prevalence rates compared with the native European population. The same group of investigators previously examined the frequencies of other vasculitides in the same geographic area, using almost identical sources of case ascertainment (2). In both instances, the authors used a capture–recapture analysis to check the validity of their sources of case ascertainment and to estimate the missing-case rate. The capture–recapture method, which has thus far been used infrequently in rheumatology, is an epidemiologic tool that allows the use of multiple samples from a single population, providing a shortcut to traditional whole population studies (3). The following 3 important conditions must be met for the capture– recapture method to have a workable validity: the population studied must be closed, like a fish pond; the likelihood of representation of the trait being examined should be equal in each sample; and capture by 1 source should not increase or decrease the likelihood of capture by another source. It is often argued that, unlike in veterinary work where the method has found wide use, the capture– recapture method generally underestimates the true disease rate when used in human medicine (4), where the fulfillment of the above 3 conditions is understandably difficult. Nevertheless, the method does give us an approximation of the underestimation. The underestimation was 20% in the study by Mahr et al published in this issue of Arthritis & Rheumatism (1) and 25% in their previous study (2), and the nonresponder rates were also quite similar (43% in the present study and 47% in the previous study). Despite the possible shortcomings of the methodology used, the fact that nearly identical methodology was used in an identical population in both studies gives us the unique opportunity to compare the frequency of BS with the frequencies of the other vasculitides in the heart of Europe. In the area near Paris, BS is several times more common than each of the 4 other vasculitides, while it is almost as common as all of them combined (Table 1). The formal designation of a disease as “orphan” in Europe requires a frequency of 1 per 2,000 inhabitants (5), and it is clear that the frequencies of BS and the other vasculitides are well within the boundaries of this definition. However, a disease can also be an “orphan” in the eyes of the physician, and we have reason to suggest that BS is considerably more orphan in this sense when compared with the other vasculitides. This contention is supported by a comparison of the space allocated to BS in standard textbooks of rheumatology with that allocated to the other vasculitides. One such well-known source (6) has 5 pages allocated to BS compared with 12 for Wegener’s granulomatosis and 7 for Churg-Strauss syndrome. In light of the new information reported by Mahr and colleagues, we hope that BS will become less “orphan” and receive more attention in textbooks, medical education and training, public policy and government, and foundationand pharmaceutical industry–supported research. Both genetic and environmental variables have been described in the etiology of BS. It has been almost Hasan Yazici, MD, Emire Seyahi, MD, Sebahattin Yurdakul, MD: University of Istanbul, Istanbul, Turkey. Address correspondence and reprint requests to Hasan Yazici, MD, Professor of Medicine, Chief, Division of Rheumatology and Department of Medicine, Cerrahpasa Medical Faculty, University of Istanbul, Safa sok. 17/7, Kadikoy, Istanbul 81310, Turkey. E-mail: hyazici@attglobal.net. Submitted for publication July 18, 2008; accepted in revised form September 17, 2008.
Read full abstract