Biases in the endometriosis literature - Illustrated by 20 years of endometriosis research in Leuven

Aim: To review the Leuven data on endometriosis to demonstrate the shifts t hat occurred over the years in diagnosis of endometriosis, classification o f women with endometriosis and thus in interpretation of results. Results: The contributions to the LUF syndrome, to non-pigmented endometriosis, to c ystic ovarian endometriosis, to deep endometriosis, to endometriosis as an immunologic disease and to the development of an animal model of endometrio sis, illustrate the persistent interest in endometriosis over 20 years. Usi ng these data it can be shown how progressively the recognition of endometr iosis caused important shifts from women who in the beginning of this perio d were classified as normal, to women who later became classified as having minimal or mild endometriosis. This was caused initially by the active sea rch for small typical lesions and later by the recognition of non-pigmented lesions as endometriosis. The second important shift was caused by the rec ognition that deep endometriosis is not only a frequent disease, but that t hese women are predominantly classified as having mild to moderate endometr iosis and even as women without endometriosis. The third shift is still ong oing, since the deep lesions reported become progressively smaller, by the "enthusiasm" of the surgeons, and by the introduction of a menstrual clinic al exam. A fourth bias in the literature concerns the diagnosis and treatme nt of cystic ovarian endometriosis. Together with these shifts in recogniti on and treatment of endometriosis, our understanding of the physiopathology of endometriosis has changed. This is illustrated by the new concepts whic h have emerged over this period. These are, the focal treatment of cystic o varian endometriosis, the concept that mild endometriosis could be a normal physiological condition and the endometriotic disease theory. Conclusion: To interpret the data of the literature we should be aware of the shifts th at have occurred in the classification of endometriosis over the past 20 ye ars, and which still can hamper the comparison of results between research groups. (C) 1998 Elsevier Science Ireland Ltd. All rights reserved.