We found that the QuickVue Influenza A+B rapid test performed moderately well, with an intermediate sensitivity and a high specificity in comparison to the CDC real-time RT-PCR assay for influenza A H1N1pdm. Furthermore, we found that test performance varied depending on how many days after symptom onset the patient presented at the clinic.
Several studies have been published on the performance of the QuickVue Influenza A+B rapid test in detecting influenza A H1N1pdm, with reported sensitivities ranging from 51 to 75% 
. Our sensitivity estimate of 64.1% falls within the range of previously reported sensitivities and is very similar to the estimate of a sensitivity of 62.7% from the Suntarattiwong et al. study in Thai children 
. In this same Nicaraguan cohort, we have previously evaluated the performance of the rapid test for seasonal influenza A in children compared to RT-PCR. In that analysis, the rapid test was found to have a sensitivity of 65.2% (95% CI 58.5, 71.4) and a specificity of 99.1% (95% CI 98.3, 99.6) which is very similar to the results found for H1N1pdm in this study 
. Likewise, in our previous study, we found a significantly lower sensitivity of the rapid test for seasonal influenza in samples collected on the day of symptom onset compared to samples collected one or two days following symptom onset.
In this study, we used separate samples for the rapid test and the RT-PCR. Although this method could result in some discordant results, we do not feel that this is a limitation of the study. Rather, in order to accurately assess the performance of the rapid test, the swab provided with the test and recommended by the manufacturer must be used. Likewise, for RT-PCR the current standard in Nicaragua and many countries is a combined nasal and throat swab. Furthermore, the rapid test swab is unusable for other diagnostic purposes once the swab is inserted into the testing solution. Thus, by adhering to the current, recommended sample collection procedures for each test respectively, this comparison allows us to assess the diagnostic accuracy of the QuickVue rapid test compared to RT-PCR in real-world conditions. In addition, the use of the two different sample types may explain the specificity of less than 100% observed in the study.
One limitation of this study is that families participating in the Nicaraguan Influenza Cohort Study are encouraged to bring their child in for medical care at the very first sign of illness, which results in a significant proportion of children presenting on the day of symptom onset in this study (27%). Since we found that rapid test performance was lower on the day of symptom onset, using this population may result in a conservative overall estimate of the rapid test performance. To address this issue, we performed sub-analyses by day of onset. For the general clinical population, the performance of the test may be closer to our estimate of a sensitivity of 72.1% and specificity of 98.4%, which we calculated for children who presented one or more days after symptom onset.
Our findings support that the QuickVue rapid test performs moderately well in the detection of influenza A H1N1pdm. Of note, we found that the test performs with low sensitivity on the day of symptom onset.