We appreciate the interest of Drs. Garner, Guillaud, and MacAulay in our work and thank them for their critical comments on our recent publication (1). Early detection of cancerous lesions is a pivotal step for reducing cervical cancer-associated mortality. Finding a method which possesses high sensitivity and specificity in screening of mass populations has been difficult for cancer researchers. False negatives are a major problem for cervical cancer testing using conventional cytologic methods and frequent retesting is required (2). Previous studies have reported that DNA ploidy testing could increase the positive prediction of precancerous and cancerous lesions of uterine cervix compared with cytology (3-5). However, there has been little data on DNA ploidy cytometry testing as a single detection method in low-resource countries where it might be used for large population screening. This was the basis for performing the randomized study under discussion. Our goal was to test the hypothesis that DNA ploidy cytometry testing would be superior to conventional cytologic screening to detect early precancerous or cancerous cervical lesions.First, we need to clarify the design of our study as to whether it was a randomized controlled trial or not. Garner et al. critically said that our study was not a randomized trial because the data analyzed in our study were pooled together from both arms. In fact, we did perform both DNA cytometry and the cytologic evaluation on all patients' specimens. This was done for ethical reasons because performing a single test on each arm would not have been approved by our institutional review boards. In Table 2 from our article, the results we reported included crossover screening. We apologize for not making this clear. However, all study processes were conducted in a randomized manner, i.e., patient's eligibility screening, randomization, blindness concealment and uncovering, intention-to-treat, and per protocol analyses. Such analyses, at first, gave an overall understanding as to the two testing methods' capability to detect possible cervical lesions. Furthermore, presenting the data in this manner did not influence the study design because this table only provided partial information about the whole study. We also performed analyses for the two separate groups (see Table 1). We felt we had made these points clear in the Materials and Methods and Results sections, but perhaps should have commented further in the Discussion.As mentioned above, the data in our article was presented inconsistently between Table 2 and in the text, which resulted in our readers' difficulty in understanding our design and data analyses. As Garner et al. pointed out, we stated that 54 cases of cancer were detected in the Results section, whereas Table 2 showed a total of 64 cases of cancer. As noted above, this is because Table 2 presented the results of crossover testing. The 54 cases stated in the Results section were the number of total cases detected by cytometry and cytology without crossover testi...