Purpose: Seroepidemiological studies on the prevalence of Hepatitis C virus (HCV) in India have been hampered by reluctance of subjects to provide blood samples for testing. We evaluated the use of saliva as an alternate specimen to blood for the detection of antibodies to HCV.
Methods: Chronic liver disease (CLD) patients attending the liver clinic were recruited for this study. A saliva and plasma sample (sample pair) was collected from each patient included in the study. Saliva samples were collected using a commercially available collection device--OmniSal. Sample pairs were tested with an in-use ELISA for the detection of antibodies to HCV (HCV-Ab), with a minor modification in the manufacturer's protocol while testing saliva. The cut-off absorbance value for declaring a sample as positive was determined by receiver operating curve (ROC) analysis. HCV-Ab positivity in saliva was compared with that in plasma as well as with viral load in plasma and infecting genotype of the virus. Sensitivity, specificity, positive and negative predictive values, and correlation coefficients were calculated using Medcalc statistical software.
Results: The optimal accuracy indices were: sensitivity-81.6%; specificity-92.5%; PPV-85.1% and NPV-90.5%. No correlation was found between salivary positivity and HCV viral load in plasma or infecting genotype.
Conclusions: The accuracy indices indicate that the assay must be optimized further before it can be recommended for routine use in epidemiological surveys for HCV-Ab.