Strengths and limitations
Randomization was one of the strengths of this study. In this way, the fluorescence colposcope group and the conventional colposcopy group could be more objectively compared in the same population. In addition, we sampled the cervix at multiple sites, regardless of the result of fluorescent dye or acetic acid and iodine staining. Previous research had shown that the strength of correlations between colposcopic impression and biopsy histology was poor, with an agreement rate of only 37%10. A recent study in China indicated the consistency between colposcopy and biopsy pathology was 59.35% with the moderate strength of kappa coefficient of 0.464 in detecting high‑grade squamous intraepithelial lesion and cervical cancer11. Multiple biopsies increased the detection of histologic HSIL, regardless of patient characteristics12. Therefore, multi-site sampling was recommended as a standard practice of colposcopic biopsy and applied in several researches13-15. And multi-site sampling also increases the sample size, which was helpful for statistical analysis. In addition to these, we quantitatively calculated the SBRs at the lesions when evaluating the results of fluorescence colposcopy, so that the results would be more objective and convenient.
This study also had some limitations or implementation challenges. Firstly, all the findings were based on our single-center data, which may cause a certain degree of bias. Secondly, neither patients nor colposcopists were blinded to the colposcopy. The colposcopy operation process of the fluorescence colposcope group was different from that of the conventional group, so the double-blind design cannot be achieved at present. Thirdly, this was a preliminary study of a new device with a relatively small sample size. Fourthly, the sampled tissue was too small to additional histological examination, and there were not enough samples for immunohistochemistry to detect the expression level of TMTP1’s potential receptor XPNPEP2. As a result, we could not evaluate the consistency of the levels of SBR and XPNPEP2 expression in the fluorescence colposcope group. In the future, the large sample multi-center clinical research should be further performed to evaluate the application potential of this fluorescence colposcope in the secondary screening of cervical cancer.