Accuracy of visual inspection performed by community health workers in cervical cancer screening

Background Cervical cancer remains the leading cause of cancer and mortality in low-resource areas with healthcare personnel shortages. Visual inspection is a low-resource alternative method of cervical cancer screening in areas with limited access to healthcare. Objectives To assess accuracy of visual inspection performed by community health workers (CHWs) and licensed providers, and the effect of provider training on visual inspection accuracy. Search strategy Five databases and four websites were queried for studies published in English up to December 31, 2015. Derivations of "cervical cancer screening" and "visual inspection" were search terms. Selection criteria Visual inspection screening studies with provider definitions, colposcopy reference standards, and accuracy data were included. Data collection and analysis A priori variables were extracted by two independent reviewers. Bivariate linear mixed-effects models were used to compare visual inspection accuracy. Main results Provider type was a significant predictor of visual inspection sensitivity (P=0.048); sensitivity was 15 percentage points higher among CHWs than physicians (P=0.014). Components of provider training were significant predictors of sensitivity and specificity. Conclusions Community-based visual inspection programs using adequately trained CHWs could reduce barriers and expand access to screening, thereby decreasing cervical cancer incidence and mortality for women at highest risk and those living in remote areas with limited access to healthcare personnel.