You are here

Article Text

Article menu

Download PDFPDF

E-Poster viewings
Pre-invasive Disease
IGCS19-0097
351 A new scoring system for colposcopy for detection cervical intraepithelial neoplasia 2+
Free
  1. V Villoslada Terrones1,2,
  2. C Acosta Cortez3,
  3. B Fora Chura4 and
  4. M Alvarez Larraondo1
  1. 1National institute of neoplastic diseases- Lima- Perú, Gynecological Surgery, Lima, Peru
  2. 2National University of Cajamarca, Data analyst- visiting professor and researcher of the School of Human Medicine, Cajamarca, Peru
  3. 3National University Agraria la Molina, statistic, lima, Peru
  4. 4Hipolito Unanue Hospital, Obstetrics and Gynecology, Tacna, Peru

Abstract

Objectives To construct scoring for colposcopic examination that can better prediction for detection CIN 2+.

Methods Six hundred and fifty patients referred for colposcopy were examined. Four variables were scored: acetowhiteness, borders, vessels, and staining time from the regression coefficients. Finally, the respective roc curves were graphed and choose the best. Hypothesis testing is done to see which curve is the best.

Results The score is summarized in table 1. The analysis showed good sensitivity and specificity see table 2. The replacement of values by the coefficients have a more useful and simple. To choose the best score, AUC graphs were made comparing the different scores graph 1.

View this table:
Abstract 351 Table 1

Scores used

View this table:
Abstract 351 Tabla 2

Parameters of AUC

Abstract 351 Figure 1
Abstract 351 Figure 1

AUC Score Total vs Aceto whitness plus borders

Conclusions Scoring needs two variables (Acetowhiteness plus borders), the best cutoff point is 7 with a Se 95%, Es 87%, PPV 94%, NPV 88%

http://dx.doi.org/10.1136/ijgc-2019-IGCS.351

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.