OBJECTIVETo evaluate the diagnostic efficiency of colloidal gold and dot ELISA rapid tests in clinical screening of influenza A virus.

METHODSThe pharyngeal swabs were collected from 297 outpatients suspected of influenza between June and October, 2009 for detection with colloid gold and dot ELISA rapid test, with real-time PCR as the golden methods. The discrepant results of colloid gold and dot ELISA methods were confirmed by sequencing, and the diagnostic efficiency of the two methods was evaluated.

RESULTSAmong the 166 samples with influenza A virus infection as confirmed by real-time PCR and sequencing, the diagnostic sensitivity of dot ELISA and colloid gold methods was 54.82% (91/166) and 4.22% (7/166), respectively. The total concordance rate with PCR was 66.67% (Kappa value of 0.35). Among the 133 samples negative for influenza A virus, the specificity of dot ELISA and colloid gold methods was 81.68% (107/131) and 98.47% (129/131), respectively, with a total concordance rate with PCR of 45.79% (Kappa value 0.02). Of the 99 H1N1 influenza samples confirmed by real-time PCR, the detection rate of dot ELISA was 67.3%, whereas that of colloid gold was 5.1%. Out of the 107 dot ELISA-positive but colloid gold-negative samples, 84 were confirmed to be influenza A virus-positive by real-time PCR and sequencing. One sample negative for dot ELISA but positive for colloid gold test was confirmed to be influenza A virus-negative. The detection rate and diagnostic concordance rate for influenza A virus by dot ELISA were significantly higher than those of colloid gold (P<0.05).

CONCLUSIONDot ELISA is better than colloid gold in influenza A virus detection and shows great prospect in clinical screening.