| Objective: To explore the availability and practicability of R-way colposcopic evaluation system in detecting CIN(cervical Intraepithelial neosplasia).Methods: A retrospective cohort study was performed as follows: Between 2014 to October 2015,a total number of 340 cases referred to colposcopy in the General Militery Hospital Of Shenyang were studied using both the R-way diagnostic system the conventional colposcopy.With reference to the 2011 edition IFCPC(International Federation of Cervical Pathology and Colposcopy)updated colposcope term,the blind method applied to R-way colposcopic diagnosis system and the traditional colposcope test(i.e.improved Reid scoring method [4](table 1),formed from modifying based on the traditional Reid scoring method),respectively.Then concluded the colposcopic diagnosis of Cervical biopsy after histopathological diagnosis as the "gold standard",the consistency between colposcopic diagnosis and pathologic diagnosis was calculated and the availability and practicability of colposcopic diagnosis were analyzed.Evaluation index mainly includes coincidence rate,sensitivity,specific,positive predictive value,negative predictive value and Youden index.Result :1.To R-way colposcopic evaluation system: the complete coincidence rate between colposcopic diagnosis and cervical histopathological diagnosis was 76.2%(259/340)and the strength of consistency with weighted Kappa value statistic was 0.448,P < 0.001;During all of the Colposcopic diagnosis,66 patients(19.4%,66/340)was overestimated and 15 cases(4.4%,15/340)was underestimated.In pathological diagnosis of normal or inflammation,LSIL,HSIL and invasive cancer,the complete coincidence rate between colposcopic diagnosis and cervical histopathological diagnosis respectively were 76.7%(214/279),58.3%(7/12),79.5%(31/39),70.0%(7/10).There were statistically significant difference between agreements in various grade lesions(X2 = 0.47,P < 0.001).Positive predictive value of high grade lesions or more was 61.5%(40/65)and the sensitivity was 81.6%(40/49),the negative predictive value of low grade lesions or less was 96.7%(266/275),whereas false positive rate was 8.6%,false negative rate was 18.4% and the specific was 91.4%(266/291).Positive predictive value of all lesions(includes LSIL and above lesions)was 43.5%(50/115)and the negative predictive value of no lesions was 95.1%(214/225).2.To improved Reid scoring method: the complete coincidence rate between colposcopic diagnosis and cervical histopathological diagnosis was 29.4%(100/340)and the strength of consistency with weighted Kappa value statistic was 0.85,P < 0.001;During all of the Colposcopic diagnosis,218 patients(64.1%,218/340)was overestimated and 22 cases(6.5%,22/340)was underestimated.In pathological diagnosis of normal or inflammation,LSIL,HSIL and invasive cancer,the complete coincidence rate between colposcopic diagnosis and cervical histopathological diagnosis respectively were 24.0%(67?279)?25.0%(3/12)?61.5%(24/39)? 60.0%(6/10).There were statistically significant difference between agreements in various grade lesions(X2 =70.093,P < 0.001).Positive predictive value of high grade lesions or more was 27.7%(31/112)and the sensitivity was 63.3%(31/49),the negative predictive value of low grade lesions or less was 92.1%(210/228),whereas false positive rate was 27.8%,false negative rate was 36.7% and the specific was 72.2%(210/291).Positive predictive value of all lesions(includes LSIL and above lesions)was 20.6%(55/267)and the negative predictive value of no lesions was 92.1%(210/228).Conclusions:1.Comparing with improved Reid scoring method,the complete coincidence rates between colposcopic diagnosis and cervical histopathological diagnosis of the R-way diagnostic system was 76.2%(259/340),which was higher than that of improved Reid scoring method 29.4%(100/340);Sensitivity of high grade lesions or more was 81.6%(40/49),which was higher than that of improved Reid scoring method 63.3%(31/49);Specific of low grade lesions or less was 91.4%(266/291),which was higher than that of improved Reid scoring method 72.2%(210/291).2.Positive predictive value of high grade lesions or more was 61.5%(40/65),which was significantly higher than that of improved Reid scoring method 27.7%(31/112);Negative predictive value of low grade lesions or less was 96.7%(266/275),which was slightly higher than that of improved Reid scoring method 92.1%(210/228).Positive predictive value of all lesions(includes LSIL and above lesions)was 43.5%(50/115),which was slightly higher than that of improved Reid scoring method 20.6%(55/267);negative predictive value of no lesions was 95.1%(214/225),which was slightly higher than that of improved Reid scoring method 91.8%(67/73).3.In short,whether consistency assessment,authentic assessment or predictive capacity in evaluation Cervical lesions,R-way colposcopic diagnosis systems are superior to improved Reid scoring method.In addition,taking into account ease of use and standardization of quality control management,R-way colposcopic evaluation system is the most desirable,relatively speaking. |
PDF Full Text Request