Publication:

Efficacy of the Swede score in prediction of high-grade lesions of cervix

Background Cervical cancer is a preventable lesion that can be identified by stepwise screening methods. Visual inspection of the cervix with acetic acid, Lugol’s iodine, pap smear, and HPV are the primary screening methods. This study aims to evaluate the efficacy of the Swede score in predicting high-grade lesions of the cervix among patients attending a colposcopy clinic.

Methods This observational cross-sectional study was performed in a colposcopy clinic under the department of obstetrics and gynaecology at CMCH. All referral patients to the colposcopy clinic were the study population. Women aged between 25-65 years was the study subject.

Results Majority (47.5%) of the patient’s Swede score was 4 and 20.1of % of patients had scores of 5-6 and only 1.3% of patients had scores ≥7%. 63 (79%) patients had low grade/normal/ cervical intraepithelial lesion (CIN) 1, 16 (20%) patients had high grade/non-invasive cancer/CIN 2 and only 1 patient had high grade/suspected invasive cancer/CIN 3. A score of 6 had a specificity of 100% for CIN 2 with a sensitivity of 30% positive and negative predictive value (PPV=100%; NPV=90.9%). Lowering the score to 5 for predicting CIN 2 improved the sensitivity at the expense of specificity (sensitivity=60%; specificity=90%; PPV= 94%; NPV=46.2%).

Conclusions Swede scoring system is consistent and reproducible, has a simple structure, and thus contributes to preventing cervical cancer. Swede score of 6 or more has 100% specificity; this scoring method is a preferred method for the treatment of high-grade CIN.