Background. HPV status correlates with spontaneous regression of cervical dysplasia. However, HPV testing is expensive and hence not readily available in low resource settings. This study aims to investigate whether demographic factors and clinical factors can correlate and hence predict spontaneous regression of cervical dysplasia.
Methods. A gynaecological practice cytology data base was used in which 142 eligible patients out of a total of 173 with positive cervical cytology tests were followed in a retrospective cohort study. The inception cohort was assessed against 7 predictor variables: age at initial positive cytology, age of onset of sexual activity, parity, smoking, oral contraceptive use, number of lifetime sexual partners and baseline grade of cervical lesion. Univariate and multivariate logistic regression for all predictor variables was performed. The Likelihood Ratio test was used to eliminate and select relevant variables. The model was validated by the use of bootstrapping.
Results. Cervical cytology was performed on 1 678 patients over the period June 2010 to April 2014. A total of 173 patients tested positive and 142 had a repeat test. Overall, 77 (54.2%) regressed and 65 (45.8%) had persistent lesions. The median age of the study was 30 years (range: 18-62). Most patients were younger than 30 (54.9%) and had fewer than 5 lifetime sexual partners (68.3%). Most had coitarche before age 18 (64.8%), did not prefer oral contraceptive use (71.1%) and did not smoke (91.6%). The majority had fewer than 2 children (86.6%), were not married (69.0%) and enjoyed tertiary education (79.6%). Age of onset of sexual activity was the only significant variable (odds ratio: 0.79; 95%CI 0.64-0.98, P=0.03). Women 30 years or older had less spontaneous regression, although non-significant (OR=0.98; 95%CI: 0.94-1.02, P=0.25). The number of lifetime sexual partners was non-significantly inversely related to spontaneous regression (OR= 0.90; 95%CI: 0.78-1.04, P=0.15).
Conclusion. Age of onset of sexual activity significantly predicts spontaneous regression of cervical dysplasia. Age at first abnormal test and number of sexual partners also predicted spontaneous regression although non-significantly.