Rotavirus, vaccine failure or diagnostic error?

Abstract

Immunochromatography (ICG) is highly used in clinical settings for rotavirus (RV) diagnosis. The specificity of the tests differs by brand type and is not 100%, therefore its use when the prevalence of the disease is low (i.e. in vaccinated children) may result in a proportion of false positive diagnoses.

In some areas, vaccine effectiveness studies or surveillance is done using ICG. Our objective was to estimate the validity of ICG test in vaccinated children, and estimate the number of false positive results in the Valencian Region of Spain, where all RV infections are diagnosed using ICG and are not confirmed by PCR.

Population based registries were used to identify all results from the RV antigen tests performed between January 2008 and June 2012 in children under 37 months. Hospitalization and vaccination status of the patients were obtained by linking different databases through a unique identification number. The Positive Predictive Value of the ICG test depending on the vaccination status of the child, hospitalization and the rotavirus season was estimated by a Bayesian model of latent classes.

Of the 48,833 tests with valid results, 9429 were done in vaccinated children, and of those 3963 (42%) during the rotavirus season. The prevalence of positive results in vaccinated varied from 2.9 to 21.4% of the tests depending on the hospitalization and seasonality. The estimated PPV also varied from 27.1 to 84.6% when stratified by these two parameters. Globally it is calculated that approximately 267 out of the 520 (51.3%) positives in vaccinated children were false positive tests.

The large percentage of false positives, due to an excessive number of tests in vaccinated and out of the RV season, if interpreted as vaccine failures, can cause a loss of confidence in the vaccine and lower the estimates of vaccine effectiveness.