Risk factors as the basis for triage in low-resource cervical cancer screening programs

Objectives: Several studies have shown impressive sensitivity and specificity of visual inspection using acetic acid wash (VIA) in detecting cervical dysplasia. This study examines how risk-factor based triage, using clinical history-taking to limit the testing to those at highest risk of disease, improves the clinical utility of VIA. Methods: The study population included 2206 woman aged 25-56 from peri-urban primary care clinics in Zimbabwe. Three risk factors variables [age, lifetime number of sexual partners and history of a sexually transmitted infection (STI)] were selected as the basis for computer-simulated patient triage. Criteria for selecting risk factors were biological relevance, programmatic feasibility, historical evidence in the scientific literature as a risk factor for cervical cancer and a significant (P<0.05) predictor of dysplasia in this study population. Predictive values for VIA were calculated on subgroups of women with combinations of the three risk factors and then compared with the predictive value of VIA calculated on the full study population. Results: The positive predictive value (PPV) among women with all three risk factors (30.2%) was 1.6 times higher than that of all study women and 4.5 times higher than women with none of the three risk factors (4.1%). Additionally, the PPV was slightly (30%) higher than that obtained from computer-simulated serial testing using VIA as the primary test followed by HPV testing (27.4%). Conclusion: Clinical history-taking could provide the basis for patient triage to increase the clinical utility of VIA. Such an approach could also serve to prioritize who gets tested first in low-resource countries with high disease prevalence that are struggling to strengthen national cervical cancer prevention programs.