Correlation of Two Colposcopic Indices for Predicting Premalignant Lesions of Cervix

Colposcopic findings to study cervical changes in reproductive age group women using various contraceptives

BACKGROUND

Colposcopy has a key role to play in see-and-treat programs for premalignant cervical lesions. The aim of the study/was to observe cervical changes with a colposcope using the Swede scoring system in fertile age group women using various contraceptives: conventional methods (barrier methods, coitus interruptus), oral contraceptives (OCPs), copper-T and bilateral tubectomy. The aim of the study was to observe and evaluate the colposcopic findings using the Swede scoring system for the diagnosis of premalignant/malignant lesions in reproductive age group women using various contraceptives.

METHODS

This was a prospective observational study, conducted among 200 women of reproductive age group using various contraceptives in a tertiary care institute in North India. PAP smear, direct visual examination, VIA (Visual Inspection with Acetic Acid) examination, colposcopic examination, and (biopsy if indicated) were done. The data were collected, and analysis was done using Microsoft Excel Office Software 2019 version 19.11 and epi info (CDC Atlanta) 7.23.1. Statistical analysis was done using percentages, mean, mode, median, standard deviation, Chi-square, Fisher's Test, and Anova Test.

RESULTS

We found positive PAP (Papanicolaou test) smears in 61.50%, positive VIA examination in 9%, and positive findings in colposcopic examination in 28.50%, Swede score of 0-3 in 100% (0-91%, 1-2%, 2-6%, and 3-1%) and positive biopsy in 9% subjects. Malignant findings were observed in 1.00% of PAP smears. Colposcopic findings were CIN 1 (cervical intraepithelial neoplasia 1) in 8.5% and CIN 2 in 0.5% subjects. Swede score was zero in 91%, 1 in 2%, 2 in 6%, and 3 in 1% of subjects. HPE (histopathological examination) was chronic cervicitis in 8.50% and mild dysplasia/CIN 1 in 0.5%. No significant statistical associations between contraceptive choice and false-positive test results or disease prevalence was found in any group except Cu-T users p = 0.0184 (especially for CIN 2; p = 0.0109 and CIN 1 more in all groups than Cu-T users). Colposcopy had sensitivity 100%, specificity 91.46% (0/0 = 0%) PPV = 5.56%, NPV = 100%, Accuracy = 91.5% for detecting mild dysplasia/CIN-non-significant (p = 0.055). Our study had mainly low-grade lesions with 100% NPV. With increase in Swede Score, sensitivity increases but at the expense of specificity but it was statistically non-significant (p = 0.055).

CONCLUSIONS

Our study may guide the rational use of colposcopy with Swede scoring for see-and-treat lesions, which is easy and with a low learning curve, as a tool for diagnosis but only in cases where indicated like unhealthy cervix because of the high rate of false-positive results. In low-grade lesions, it is highly useful to rule out the disease.

Association of Swede Score and 2011 IFCPC Nomenclature in Women with Abnormal Cytology

OBJECTIVE

 To assess the association between two colposcopic indices, the Swede score and the 2011 International Federation of Cervical Pathology and Colposcopy (IFCPC) Nomenclature as well as to determine the efficacy of the Swede score with cutoffs of 7 and 8.

METHODS

 In the present cross-sectional pilot study, 34 women who had at least 1 colposcopy-directed biopsy due to abnormal cytology were enrolled. The colposcopic findings were scored by both the Swede score and the 2011 IFCPC Nomenclature and were compared with each other. The Kappa coefficient and the McNemar test were used. Accuracy, sensitivity, specificity, and positive and negative predictive values (NPV and PPV, respectively) were calculated, as well as the effectiveness with cutoffs of 7 and 8 in identifying cervical intraepithelial neoplasm (CIN) 2+ when using the Swede score.

RESULTS

 The correlation between the 2 colposcopic indices was 79.41%. The Kappa coefficient and the McNemar p-value were 0.55 and 0.37, respectively. The IFCPC Nomenclature had sensitivity, specificity, accuracy, PPV, and NPV of 85.71, 55.00, 67.64, 57.14, and 84.61%, respectively. The Swede score had sensitivity, specificity, accuracy, PPV, and NPV of 100, 63.15, 79.41, 68.18, and 100%, respectively. A Swede score cutoff of 7 for CIN 2+ detection had a specificity of 94.73%, while with a cutoff of 8 it increased to 100%. The sensitivity for both values was 60%. The PPV and NPV for cutoffs of 7 and 8 were 90 and 75 and 100 and 76%, respectively.

CONCLUSION

 Although both colposcopic indices have good reproducibility, the Swede score showed greater accuracy, sensitivity, and specificity in identifying CIN 2 + , especially when using a cutoff of 8.

The Diagnostic Efficacy of Swede Score for Prediction of Pre-invasive Cervical Lesions: A Prospective Hospital-Based Study

Context

The accuracy of colposcopy has recently been called into question particularly with regard to determining the site requiring biopsy. The technique of colposcopy is largely operator dependent, and the agreement between colposcopists is less reliable than once presumed. In an attempt to standardize colposcopy a new scoring system, the Swede score has been devised, which includes lesion size as a variable to be scored in addition to the 4 variables found in the modified Reids Colposcopic Index (RCI).

Aim

To assess the diagnostic accuracy of Swede score for pre-invasive cervical lesion.

Setting and Design

A cross-sectional study in a tertiary care centre.

Method

Swede score was calculated for assessment of pre-invasive cervical lesions on patients undergoing colposcopy who were suspected with pre-invasive cervical lesion. Cervical biopsy was taken if modified RCI ≥ 3 or Swede score ≥ 5. Histopathology report of the cervical biopsy was taken as gold standard.

Results

Swede scores of 5 or more had sensitivity, specificity, positive and negative predictive values of 94.9%, 88.4%, 75.5% and 92.9% respectively.

Conclusion

The Swede score by just incorporating one additional variable that is size of the lesion, showed better correlation with histopathology.