tubo-ovarian inguinal herniation. However, irrespective of the actual etiology, an inguinal ovary diagnosed during child- hood presents as a hernia and warrants a herniorrhaphy (5). Consequently, the authors’ contention that a history of an inguinal operation in their patient raises the suspicion of a herniation of the left ovary (as opposed to another etiol- ogy preventing ‘‘the ascent of the ovary’’) is not validated. While clomiphene citrate (CC) use has facilitated the diag- nosis of an abdominal ‘‘undescended’’ abdominal ovary on magnetic resonance imaging (2), we do not find this diagnos- tic tool to be necessary for the diagnosis of the inguinal ovary, where the ultrasound findings of different sized follicles can- not be mistaken. At least theoretically, administration of CC in the case of an inguinal ovary is not innocuous. Tony Bazi, M.D. Muhieddine Seoud, M.D. Department of Obstetrics and Gynecology American University of Beirut Beirut, Lebanon Samir Akel, M.D. Department of Surgery American University of Beirut Beirut, Lebanon September 20, 2006 REFERENCES 1. Idil M, Ozdemir BG, Ocal P, Cepni I, Erturk S, Erguney S. Detection of an inguinal ovary at controlled ovarian stimulation that was successfully treated by repositioning. Fertil Steril 2006;85:1822.e9–11. 2. Ombelet W, Grieten M, DeNeubourg P, Verswijvel G, Buekenhout L, Hinoul P, et al. Undescended ovary and unicornuate uterus simplified di- agnosis by the use of clomiphene citrate ovarian stimulation and magnetic resonance imaging (MRI). Hum Reprod 2003;18:858–62. 3. Hinckley MD, Milki AA. Percutaneous oocyte retrieval from an inguinal ovary. Fertil Steril 2003;80:445–6. 4. Mayer V, Templeton FG. Inguinal ectopia of the ovary and fallopian tube review of the literature and report of the case of an infant. Arch Surg 1941;43:397–408. 5. Bazi T, Berjawi G, Seoud M. Inguinal ovaries associated with Mullerian agenesis: case report and review. Fertil Steril 2006;85:1510.e5–8. doi:10.1016/j.fertnstert.2007.03.012 Verification bias? To the Editor: Although the hypothesis proposed by Keltz et al. that chla- mydia antibody titres may aid in the detection of tubal disease is plausible, we feel that the conclusion that chlamydia serol- ogy is an inexpensive, noninvasive test that matches or sur- passes the predictive value of most standard infertility tests is premature (1). The study of the test characteristics of a new diagnostic test is often hampered by the logistics of a difficult study design. ‘‘In the ideal study of a test, each patient in a representative sample undergoes both the index test and the gold standard pro- cedure’’ (2). In this case, no uniform gold standard was used to calculate sensitivity, specificity, or negative and positive pre- dictive values for chlamydia serology as a screening test; that is, not all patients underwent laparoscopy. Given the lack of consistency in the use of the gold standard test, it is difficult to evaluate the true performance of any screening test. In addition, data should optimally be presented so that 2 Â2 tables can be recreated to assess each test’s performance and to confirm reported sensitivity, specificity, and predictive values. For example, in Table 1 of Keltz et al., it is unclear how 18 patients had laparoscopically confirmed bilateral tubal blockage but only 15 had tubal disease. We were also unable able to recreate the data tables that assess the per- formance of hydrosalpingography (HSG) compared with lap- aroscopy. In addition, the authors did not present the predictive value of HSG. Moreover, confidence intervals (CIs) should be included to assess the precision of each point estimate. We were able to recreate the serology table, but when we calculated the CI for the specificity of serology (82.4, 98.6), it was wide and included the test performance of HSG, which does not support the authors’ conclusion that specificity for serology surpasses that of HSG. Finally, while the authors state that antibody testing is in- expensive, it is unclear what added value this additional test- ing provides based on the presented data. It is premature to conclude that serology ‘‘matches or surpasses’’ the predictive value of standard infertility tests. We caution the recommen- dation to consider chlamydia serology as an adjunct screen- ing test to help direct the management of infertile patients in the absence of more compelling evidence. At this time, given the broad specificity and the continued need for HSG, it is unclear what benefit serology screening provides given the additional cost. Suleena Kansal Kalra, M.D. Beata Seeber, M.D. Sindhu Srinivas, M.D. Kurt Barnhart, M.D., M.S.C.E. Hospital of University of Pennsylvania, Philadelphia, Pennsylvania September 25, 2006 REFERENCES 1. Keltz MD, Gera PS, Moustakis M. Chlamydia serology screening in infer- tility patients. Fertil Steril 2006;85:752–4. 2. Sox HC, Blatt MA, Higgins MC, Marton KI. Medical decision making. Woburn, MA: Butterworth-Heinemann, 1988. doi:10.1016/j.fertnstert.2007.06.001 Re: Gel instillation sonohysterography: first experience with a new technique To the Editor: The article by Exalto et al. (1) on gel instillation sonogra- phy (GIS) made an interesting read. It seems that a minor 536 Letters to the Editor Vol. 88, No. 2, August 2007