The Swede Score Colposcopic Evaluation in Prediction of High-Grade Lesions in Human Immunodeficiency Virus-Infected Patients with Abnormal Cervical Cytology

Abstract

Objectives: To evaluate the performance of the Swede score colposcopic scoring system at the different cut-off scores to predict high-grade lesions (HGL) in human immunodeficiency virus (HIV)-infected patient with abnormal cervical cytology.

Materials and Methods: A total of 80 HIV-infected patients undergoing colposcopic evaluation at Chonburi hospital, Thailand, due to their abnormal cervical cytological screening results, were included. All participants were evaluated with colposcopy and the findings were scored according to the Swede score system. Colposcopic directed cervical biopsy was performed in all cases. Then the final pathological diagnosis was compared with the recorded score. The performances of the Swede score for the prediction of HGL was determined by estimating the sensitivity, specificity, positive and negative predictive values of each cut-off point of total Swede score.

Results: The prevalence of HGL was 23.8%. All cases of HGL had a Swede score of 4 or above; therefore, sensitivity and negative predictive value for HGL at the cut-off score of 4 was 100%, and 100%, respectively. The maximum positive predictive value was only 47.8% at the cut-off score of 7 with the specificity of 80.3%. The specificity values at the cut-off scores of 8, 9, 10 were 83.6%, 88.5%, and 95.1%, respectively.

Conclusion: Swede Score was a good colposcopic scoring system which could be used as a screening tool to accurately exclude HGL in HIV-infected patients with abnormal cervical cytology when the total score was below 4. The specificity in HIV-infected patients was significantly lower than of that in non-HIV-infected patients; therefore, it required pathologic confirmation before treatment.